What Mathematics Can Do For You

Yoshikazu Giga · Toshiyuki Kobayashi
Editors
What
Mathematics
Can Do for You
Essays and Tips from Japanese Industry
Leaders
What Mathematics Can Do for You
Yoshikazu Giga • Toshiyuki Kobayashi
Editors
What Mathematics
Can Do for You
Essays and Tips from Japanese
Industry Leaders
123
Editors
Yoshikazu Giga Toshiyuki Kobayashi
Professor Professor
Graduate School of Mathematical Sciences Graduate School of Mathematical Sciences
The University of Tokyo The University of Tokyo
and
Principal Investigator
Kavli IPMU
ISBN 978-4-431-54345-9 ISBN 978-4-431-54346-6 (eBook)

DOI 10.1007/978-4-431-54346-6
Springer Tokyo Heidelberg New York Dordrecht London
Library of Congress Control Number: 2013936747
© Springer Japan 2013

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection
with reviews or scholarly analysis or material supplied specifically for the purpose of being entered
and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of
this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from Springer.
Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations
are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Preface
Japan is a tiny country that occupies only 0.25% of the world’s total land area.
However, this small country is the world’s third largest in economy: Japanese GDP
is roughly equivalent to the sum of any two major countries in Europe as of 2012.
This book is a first attempt to ask business leaders of top Japanese companies
for their thoughts on mathematics. The topics range from mathematical problems
in specific areas (e.g., exploration of natural resources, communication networks,
and finance, etc.) to mathematics literacy that would help a leader who has to
weigh many different issues, and even to mathematical thinking in connection to
quality control and a long-term strategy. We express our deep gratitude to these
business leaders who shared their thoughts on mathematics with us in spite of their
extremely busy schedules. The reader may notice that every article—sometimes
even the choice of vocabulary—reflects the authors’ ways of life and thinking.
This book is an enlarged English edition of the Japanese version, What Mathe-
matics Can Do for You—Essays and Tips from Japanese Industry Leaders. For this
edition we have invited three mathematicians who have been trying to expand and
strengthen the interaction between mathematics and industry.
Mathematics asks nothing in return. The role of mathematics is often invisible
when it is applied effectively and smoothly in science and technology, and math-
ematical strategy is usually hidden when it is used properly and successfully. It
is a pleasant surprise to us that the leaders in global companies appreciate this
invisible feature of mathematics. The editors hope this book will give the reader
an opportunity to notice something hidden but important.
This book owes much to the invaluable support of Mr. Toru Yosano, Senior
Advisor of BNP Paribas Securities (Japan) Limited, and of Springer Japan, from an
earlier stage of the project. We also thank our colleague Professor Shigeo Kusuoka
for his kind suggestions.
Tokyo, Japan Yoshikazu Giga

Tokyo, Japan Toshiyuki Kobayashi
v
Contents
A Dialogue Between a Mathematician and Toyota’s Chairman:

Think, Think, and Think Again . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 1
Fujio Cho and Toshiyuki Kobayashi
BNP Paribas/Tokyo: Reasons Why Mathematics Is Important
to Our Company . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 7
Yusuke Yasuda
Takeda Pharmaceutical Co., Ltd.: Are Numbers Useful?
Acknowledging the Contribution of Mathematical Modeling
to Corporate Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 17
Yasuchika Hasegawa
Nippon Telegraph and Telephone Corporation (NTT):
Mathematics Drives the Economy .. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 23
Norio Wada
Nomura Holdings, Inc.: The Role of Mathematics in Finance:
Applied Mathematics and Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 29
Kenichi Watanabe
HORIBA, Ltd.: Mathematics Is the Starting Point
of Corporate Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 35
Atsushi Horiba
Railway Technical Research Institute (RTRI): Mathematics
Supports Development of Railway System Technology ... . . . . . . . . . . . . . . . . . . . 45
Eisuke Masada
Japan Oil, Gas and Metals National Corporation (JOGMEC):
The Role of Mathematics in the Petroleum and Natural Gas
Exploitation Industry in Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 53
Hirobumi Kawano
vii
viii Contents
Iwane Laboratories, Ltd.: Mathematics in Our Company:

What Does It Describe? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 65
Waro Iwane
Mathematics and I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 73
Kaoru Yosano
Mathematics for Industry: Principle, Reality and Practice,
from the Point of View of a Mathematician . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 77
Masahiro Yamamoto
Importance and Unpredictable Effectiveness of Mathematics
in the Real World and for Industry . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 101
Masato Wakayama
Mathematics for Business and Business Leaders Based
on Mathematical Intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 123
Hiroshi Fujita
A Dialogue Between a Mathematician
and Toyota’s Chairman:
Think, Think, and Think Again
Fujio Cho Toshiyuki Kobayashi

Chairman of the Board, Professor,
Toyota Motor Corporation Graduate School of Mathematical Sciences,
The University of Tokyo and Kavli IPMU
What Lies Beneath
Kobayashi (K): I believe that there are a variety of things that are fundamental but
rarely perceived, that are important but too obvious to attract our attention. Do you
think the value of honesty for a company is such a thing? I surmise that a sound
company does not have to think about “the value of honesty” so often because
people there just take it as a matter of course. On the other hand, dishonesty—
needless to say fraud, but also pretending to know everything without performing
investigations, which I also consider a kind of dishonesty—seems to result in a long-
term loss even though it may produce a large short-term gain.
Today I would like to talk with you, by highlighting things that we are usually
unaware of. Of course, there are two opposite cases for these things, namely, because
they are so natural that we take them for granted, or simply because they are not
important.
Mathematics might be the former case. The smoother mathematics is functioning
within society, the more people are unaware of it. Mathematics is thought of as
fundamental in science and technology. Nevertheless it might go well for the time
being even if we ignore mathematics and forget thinking about reason and logic.
But this would not be the right way if we hope to achieve long-term success.
Y. Giga and T. Kobayashi (eds.), What Mathematics Can Do for You: Essays and Tips 1
from Japanese Industry Leaders, DOI 10.1007/978-4-431-54346-6 1,
2 F. Cho and T. Kobayashi
If mathematics could be useful in the real world, how? Since I am a mathemati-

cian, I am afraid that my answer to this question would be biased. Instead, I would
like to hear opinions of a Japanese business leader.
Cho (C): How clear your explanation is! Since I was a law student at the
university, I did not take math classes after finishing my sophomore year. So, my
knowledge of mathematics is very limited, especially when my conversation partner
is a professional. However, I believe mathematics is the science that has the most
logical and reasonable structure.
(Mathematical) Reason and Sentiment
C: Reason is antithetical to sentiment. It is very important to use rationality while

organizing information. I think mathematics is the best practice for developing our
reason.
Math Is Fun
C: Until I was a sophomore in high school, I was more interested in judo and kendo
than classroom studies. I studied math very hard over the next 2 years. I had been
told by my father that the family’s economic circumstances did not allow me to go to
a private university. So I had to study eight subjects, including Math I and Math II.
But I liked math because math valued logical thinking.
My father also graduated from a law school, and he was a very logical-minded
man. When I was in junior high and high school, he used to talk me down logically
by asking the reason for my argument. When I said something off the top of my
head, I was often scolded, with him saying, “You have to think more logically.”
When I studied math, I realized how my father’s advice was valuable. In the second
year of high school, I decided to major in the humanities in college, but math was
still fun to me. Looking back, the relationship with my father had an influence on
my bent toward mathematics.
Mathematics and “Kaizen”
C: After graduating from the university, I started working for Toyota. Six years later,
I was told to do “Kaizen” (improvement of production processes), although what I
had majored in was law, not engineering. Kaizen is an engineering person’s job, I
thought. I was embarrassed and tried to persuade my boss to reconsider his decision,
but in vain. I have now been doing Kaizen for 15 years. Looking back on these 15
years, I find I learned many things, mainly two lessons.
Toyota Chairman vs. a Mathematician 3
The first lesson I learned was the fact that production processes consisted of many
numbers. It seems there is something related to what Prof. Kobayashi mentioned
previously with regard to the term “honesty”: If people in a factory do their jobs
based on numbers, there can be little room for cheating and fudging. To be based on
numbers is the same thing as to be based on honesty. Production planning—which
sort and how many vehicles will be produced, how much sales are expected, how
much the cost price will be, etc.—is totally based on numbers. Without numbers,
we can say nothing but subjective remarks like “That vehicle looks cool” or “This
one provides a comfortable ride.” When we have to develop a new car, we cannot
depend on such empty words. What we need is information that is expressed with
numbers, such as bulk weight, noise level, vibration level, etc. By “numbers,” I
am not talking about technical calculations that are conducted in laboratories. I am
talking about simple arithmetic, such as addition, subtraction, multiplication and
division. Numbers and arithmetic are what Toyota’s 300,000 employees share as the
fundamental communication tool.
Ask “Why” Five Times
C: The first lesson I learned from Kaizen is the importance of numbers, which are
objective tools for us rather than abstract ideas. On the other hand, the second lesson
I learned has something more fundamentally mathematical. Doing mathematics and
being logical seem to me to be rather the same thing.
“All things have their causes.” Everyone knows that. However, people in our
company do not stop here, but keep thinking that “All things have their causes, and
all causes also have their real causes.” They should track back from causes to causes
until they find real causes. To do so, they should ask themselves the question “why”
at least five times.
They should go to factories themselves, observe what’s going on there, and
use their own hands to measure and count. They should not make their tasks
self-contained. When there is a good business, there is a sound logic behind it.
Everything derives from logic and numbers, and nothing in the world will happen by
chance. So, I was always tracking back problems until I found effective measures.
I kept asking questions like “Why does the cost price keep rising?” “Why did
this machine break down?” and “Why are there so many defective parts?” After
thoroughly asking myself such questions, I was able to understand the real causes
of a particular problem.
One day in a factory, I encountered a strange case. The blade in a certain machine,
which was supposed to last at least 1 month, was dead in the first week. I guessed
the supplier of the blade might have mixed in the wrong ingredients, but it was not
the case. Then, why? What made the blade break so quickly? That the machine’s
blade broke once a week meant that one could observe the instance of a break if one
kept watching it for a week! We worked in rotation and kept watching it. We noticed
that the blade broke because iron scraps wound around it and exerted a strong force
on it. Then, why did the scraps wind around the blade? We found many plausible
causes for this phenomenon. The solution we finally found was to attach a cutting
device to the machine, so that the produced scraps were made too short to wind
around the blade.
These experiences made me develop a habit of seeing things logically. If my
father were still living today, he would be a little proud of me.
K: I am delighted to know your “5-Whys” principle in the Toyota Company! As
a mathematician, I think that such a spirit is crucial also in creating a new theory of
mathematics. As an educator, this is what I am trying to emphasize in classrooms.
Though I usually teach graduate students majoring in mathematics, I like giving
elementary courses for undergraduates, in particular, for students majoring in the
humanities, law, or economics, when my schedule permits it. In these courses, I
advise students not to memorize individual mathematical formulas because what
is crucial is to understand them systematically. I feel it more important to make
an effort to understand the logic and the rationale and to see through something
universal. For this I suggest that they cultivate the habit during my semester course
in the following way: “Do not avoid the things that you do not understand,” “Try to
clarify, capture and pin down by your own language what you do not understand”
and “Keep thinking about it again and again.”
C: That’s a good idea!
K: Toyota’s 5-Whys seems to me to share a quite similar basis. A blind
acceptance of knowledge is a facile solution but may result in an obstruction to
making further progress. We often realize that we have understood nothing. Starting
from this, we work hard to try to understand something better, and eventually we
may reach the painful realization that there are even more things that we do not un-
derstand! I think this is tough but important training for thinking with concentration.
I would like to cultivate students by giving them the opportunities in my classes to
deeply contemplate rather than giving up quickly or just memorizing. In fact, this is
my primary motivation for giving undergraduate courses in mathematics.
C: I see your point. When studying math, we have to think, not memorize.
K: Exactly. Further, also in economics, I would be afraid an attitude to memorize
“laws” and “formulas” or even the “notions” would end up misusing them in the
real world. This would be dangerous. Instead there is another discipline not to
memorize but to analyze what is crucial by thinking and thinking about what is more
fundamental. It takes an enormous amount of time. In my opinion, deep thinking for
the universal truth is the essence of mathematics.
C: As a high school student, I was trying to solve trigonometry problems. They
were very difficult ones, and I did not even have an idea where to start. I had been
thinking about one of these problems for a whole day and finally found the answer.
After solving the first problem, however, I could solve the remaining five to seven
problems almost instantly. It seemed something inside me had changed after taking a
whole day to solve the first problem. It was unforgettably impressive to me. I learned
from this experience the importance of thinking things through.
K: The whole day that you spent by thinking with “I-don’t-sees” must be a really
special time to bring you the breakthrough afterwards. As a mathematician, I always
Toyota Chairman vs. a Mathematician 5
have many I-don’t-sees while attacking difficult problems that nobody in the world
knows how to solve, or trying to develop a new mathematical theory without existing
methods as if I were in complete darkness. No matter how uncomfortable I feel the
I-don’t-sees, I am still optimistic and want to grapple with them face to face,
C: It seems that mathematical research embraces quite many “whys”.
K: That’s the point! By that very reason, your 5-Whys motto is impressive to me.
It is a clear, nice and encouraging message that will be retained in the memory.
Company Profile of Toyota
Toyota Motor Corporation was founded in 1937. The company’s headquarters is

located in Toyota City, Aichi prefecture. It manufactures and sells conventional
passenger cars such as the Toyota Corolla, as well as hybrid and other types
of eco-friendly vehicles. Toyota has overseas manufacturing companies in 27
countries/regions, and its vehicles are sold in more than 160 countries/regions.
Toyota is listed on the New York and London stock exchanges, as well as on
the major markets in Japan. In the fourth quarter of FY2011, it made a net profit of
283.5 billion yen by selling 735,000 vehicles worldwide. As of March 2012, about
320,000 employees were working for the company.
A Brief Biography of Fujio Cho, Toyota
Fujio Cho was born in 1937. Following graduation from the University of Tokyo
with a Bachelor’s Degree in Law, he joined the Toyota Motor Corporation in 1960.
In 1988, he became a company director and was appointed CEO of Toyota Motor
Manufacturing, USA. After holding posts as executive director, senior executive
director and executive vice-president, he became the company CEO in 1999. He
was appointed as the company chairman in 2006 and also served as director of Sony
Corporation and Central Japan Railway Company. He is also the chairman of the
Japan Sports Association.
His awards include the Grand Cordon of the Order of the Rising Sun and the
Medal of Honor with Blue Ribbon.
About the University of Tokyo
The University of Tokyo was founded in 1877 by succeeding older schools from the
Edo government, applying the education systems and curriculums introduced from
the West after the restoration of Imperial power in 1868.
Among prominent people graduated from Japan’s oldest university, there are
seven Nobel laureates (Yasunari Kawabata, Reona Ezaki, Eisaku Sato, Kenzaburo
Oe, Masatoshi Koshiba, Yoichiro Nambu and Ei-ichi Negishi) and a Fields Medalist
(Kunihiko Kodaira).
Known by his contributions to education and research in Japan is Dairoku

Kikuchi, one of the university’s founding professors. After working as President of
the university, he became President of Kyoto Imperial University (the predecessor
of Kyoto University), Minister of Education and the first president of the Science
Research Institute of Japan (RIKEN).
According to Thomson Reuters’ World University Rankings, the University of
Tokyo was ranked 11th in 2010. There are 30,000 students in the graduate and
undergraduate schools combined, and 15% of the graduates consist of international
students.
A Brief Biography of Toshiyuki Kobayashi, Mathematician
Toshiyuki Kobayashi, a mathematician, was born in 1962. When he was 25, he

proved a necessary and sufficient condition for the Calabi–Markus phenomenon,
which led him to create a new theory of discontinuous groups beyond the framework
of Riemannian geometry.
Further, he has successively created his original theories of mathematics, includ-
ing the “theory of discrete decomposable branching laws” on infinite-dimensional
spaces and “visible actions on complex manifolds” aiming for the unification of
multiplicity-free representation, to name a few.
His academic honors include the Spring Prize (the Mathematical Society of
Japan), Osaka Science Prize, JSPS Prize (the Japan Society for the Promotion of
Science), Sackler Distinguished Lecturer (Israel) and Humboldt Research Award
(Germany).
He is a professor of the Graduate School of Mathematical Sciences, the Univer-
sity of Tokyo and also is a Principal Investigator of Kavli Institute for the Physics
and Mathematics of the Universe (IPMU).
BNP Paribas/Tokyo
Yusuke Yasuda
Former Chief Country Representative
BNP Paribas/Tokyo
Reasons Why Mathematics Is Important to Our Company
When trying to demonstrate the necessity of mathematics in finance, I usually begin

by explaining the Black–Scholes model, as an application of stochastic differential
equations. But in this case, I would like to change the perspective a little and talk
to you about mathematics as a tool for thought, on how effective a tool it can be, in
both English and Japanese, and also for people who work in finance.
Therefore, I would like to talk about the basic financial concept of “interest rate”
rather than about something highly advanced.
Definitions
If I invest X(0) amount of money at time 0, and collect it at time τ , I will receive
X(τ ) determined by the following equation:
X(τ ) = (1 + rτ · τ )X(0) (1)
where rτ is the spot interest rate for the period [0, τ ]. In addition, X(0) is the
principal which was initially invested, and ΔX ≡ X(τ ) − X(0) is the interest. In
actual business, τ can be a period of 1 day, 1 week, 1 month, 2 months, or even up
to 12 months. The spot interest rate used for the period of 1 day is referred to as
the overnight rate; that of the 1 week period is referred to as the 1-week rate; and
8 Y. Yasuda
so on. Also, I can reserve an interest rate, for example, at a future t, in which case
I promise to invest the amount of X(t) at time t and in return for this investment
at time of t + τ I will receive the amount of money determined by the following
equation:
X(t + τ ) = (1 + Rt→t+τ · τ )X(t). (2)
The interest rate Rt→t+τ for the future time period [t,t + τ ] can be called a “reserved
interest rate”.1 This reserved interest rate at t = 0 is the interest rate for the period
[0, τ ], therefore should equal to rτ , that is:
R0→τ = rτ . (3)
If I decide to invest the principal X(0) for the period [0, τ ] at the spot interest rate
R0→τ (= rτ ), and furthermore decide to reinvest the principal and interest at time τ
for the next period [τ , 2τ ] at the reserved interest rate of Rτ →2τ etc., I will receive
the aggregated interest and principal at time nτ (n is an integer greater than one) in
accordance with the following equation:
X(nτ ) = (1 + R0→τ · τ )(1 + Rτ →2τ · τ ) · · · (1 + R(n−1)τ →nτ · τ )X(0). (4)
Of course, τ can be e.g., 0.5 years, and therefore the amount after 0.5 years, 1 year,
and 1.5 years will be:
X(0.5y) = (1 + R0→0.5y · 0.5y)X(0), (4)
X(1y) = (1 + R0→0.5y · 0.5y)(1 + R0.5y→1y · 0.5y)X(0), and (4)
X(1.5y) = (1 + R0→0.5y · 0.5y)(1 + R0.5y→1y · 0.5y)(1 + R1y→1.5y · 0.5y)X(0). (4)
Usually, the interest rate will be a positive number greater than zero, and therefore
X(0) < X(0.5y) < X(1y) < X(1.5y) < · · · < X(nτ ) < · · · . (5)
Thus the amount of money will continue to increase as the original principal and
interest will be reinvested over time. In other words, the value of money will
continue to increase with time. The Eq. (4) gives other important information,
namely, that X(0) and X(nτ ) are in proportion:
X(0) = X(nτ ) · DFnτ . (6)
1 Inmany finance-related books, this is also referred to as forward rate or interest rate futures, but
for the sake of emphasizing the interest rate which is reserved now, I would like to refer to it as
“reserved interest rate”.
BNP Paribas/Tokyo 9
DFnτ is referred to as the discount factor at time nτ . This equation means that the
money at time nτ should be discounted by DFnτ to evaluate its present value. By (4)
and (6), DFnτ can be expressed as follows:
1
DFnτ ≡ . (7)
(1 + R0→τ · τ )(1 + Rτ →2τ · τ ) · · · (1 + R(n−1)τ →nτ · τ )
More practically, the discount factor at each time 0.5, 1, 1.5 years etc. can be
expressed as follows:
1
DF0.5y ≡ , (7)
(1 + R0→0.5y · 0.5y)
1
DF1y ≡ , (7)
(1 + R0→0.5y · 0.5y)(1 + R0.5y→1y · 0.5y)
1
DF1.5y ≡ , · · · . (7)
(1 + R0→0.5y · 0.5y)(1 + R0.5y→1y · 0.5y)(1 + R1y→1.5y · 0.5y)
This concept of the discount factor is very important in finance. For example, when
comparing two different investment schemes, in order to decide which is more
profitable, one cannot decide this by merely comparing the money at the end, but
one needs to match the time of investment. Therefore, we can use the discount factor
and calculate the present value of each investment. By (5) and (6), we can see:
1 > DF0.5y > DF1y > DF1.5y > · · · > DFny > · · · > 0. (8)
This is a universal rule in the world of finance. Please take note of this, as I will
explain in the latter part of the chapter.
Bootstrapping Method
In the case of investing longer than 1 year, it is common to receive interest in regular
intervals. For investments longer than 1 year, it is common to receive interest every
6 months using the following equation:
c1
X(0). (9)
2
The reason c1 is divided by 2 is because the amount of interest for 1 year is c1 X(0),
with half of this amount being paid in 6 months, and the remaining half paid in 1
year. The yield (in this case c1 ) is called annual yield. The amount of money being
received during the investment period is:
10 Y. Yasuda
c1
the interest in 6 months: X(0) and
2
c1
the principal and the interest in 1 year: X(0) + X(0).
2
Using the discount factor of DF0.5y and DF1y , we have
c1 c1
X(0)DF0.5y + X(0) + X(0) DF1y . (10)
2 2
This is the present value of the amount of money which will be received in the
future. In the financial world, this is regarded as the original investment. If c1 is too
small, the investment is not too attractive and no one will invest in it, and if it is
too large, everyone will rush to invest. Therefore there always exists an equilibrium
point, and it is fair to say that the equilibrium point is determined where the present
value of the amount received in the future is equivalent to the initial investment.
This assumption gives us the following equation:
c1 c1
X(0) = X(0)DF0.5y + X(0) + X(0) DF1y .
2 2
If both sides are divided by X(0), this equation becomes:
c1 c1
1= DF0.5y + 1 + DF1y . (11)
2 2
Referring back to the Eqs. (7) and (7) , we see that DF0.5y can be derived using
the spot interest rate R0→0.5y (= r0.5y ), and if the reserved interest rate R0.5y→1y is
determined, DF1y will be fixed. Using this in (11), we can calculate c1 . However, if
the value of c1 is fixed, we can substitute c1 and DF0.5y into (11) and calculate DF1y ,
and therefore the reserved interest rate R0.5y→1y is determined. In this way, we can
see that c1 and R0.5y→1y determine each other uniquely: if either of them is fixed, we
can compute the other.
In the next example, we will consider an investment of 1.5 years. If we assume
that the annual yield is c1.5 , in which case we receive interest every 6 months for the
period of 1.5 years, the amount of money received in this period will be:
c1.5
the interest in 6 months: X(0),
2
c1.5
the interest in 1 year: X(0), and
2
c1.5
the principal and the interest in 1.5 years: X(0) + X(0).
2
By the same method as before, it can be determined that
c1.5 c1.5 c1.5
1= DF0.5y + DF1y + 1 + DF1.5y . (12)
2 2 2
From this equation, we can see that c1.5 and R1y→1.5y also uniquely determine each
other. Similarly, we can determine the relationship between c2 and R1.5y→2y , and
between c2.5 and R2y→2.5y . This method is called the “bootstrapping method” since
BNP Paribas/Tokyo 11
we tie the relationship between the pairs of variables from the bottom to the top of
the boot.
The Market Convention Equation
In the financial industry, the equation for calculating investments longer than 1 year
comes from market convention. Let n be the number of years (n is an integer greater
than two), and let cn be the annual yield, with the assumption that you receive
interest every 6 months for the period of n years. In this case, if pn is the annual yield
if you receive interest every year for the same period, the general market convention
for calculating pn is given by:
cn 2
1 + pn = 1 + . (13)
2
This equation has been used for many years and has become the norm. People not
familiar with finance may think that interest received every year is simply double the
interest received every 6 months (i.e. 2× c2n ), but in the financial world, if the amount
of money received is the same, receiving it earlier is always beneficial. Therefore
the amount of 2 × c2n should be smaller than that of pn , in other words, cn < pn . If we
take (13) and solve for the right-hand side, we can easily confirm this. This equation
is an oversimplified version and therefore problematic. To confirm this, we should
try to solve for cn and pn using the bootstrapping method.
First, the annual yield cn in which you receive the interest every 6 months gives
the following equation:
cn cn cn cn cn
1= DF0.5y + DF1y + DF1.5y + · · · + DF(n−0.5)y + 1 + DFny . (14)
2 2 2 2 2
Solving for cn gives:
2(1 − DFny)
cn = . (15)
DF0.5y + DF1y + DF1.5y + · · · + DF(n−0.5)y + DFny
Secondly, the annual yield pn gives the following equation:
1 = pn DF1y + pn DF2y + · · · + pn DF(n−1)y + (1 + pn)DFny . (16)
Therefore pn is given by:
1 − DFny
pn = . (17)
DF1y + DF2y + · · · + DF(n−1)y + DFny
If we assume (13) holds, by substitution in (15) and (17) we obtain:

12 Y. Yasuda
2
1+DF1y +· · ·+DF(n−2)y +DF(n−1)y 1+DF0.5y +· · ·+DF(n−1)y +DF(n−0.5)y
= .
DF1y +DF2y +· · ·+DF(n−1)y +DFny DF0.5y +DF1y +· · ·+DF(n−0.5)y +DFny
(18)
After a straightforward calculation, we find that (18) can be true only in the
following case:
1 DF0.5y DF(n−0.5)y
= = ··· = . (19)
DF0.5y DF1y DFny
And by substituting the equations of discount factors (7) , (7) , (7), . . . , we find
that
R0→0.5y = R0.5y→1y = · · · = R(n−0.5)y→ny . (20)
This means that the reserved interest rates will all be the same. However, this
contradicts what happens in the real world. Therefore we conclude that Eq. (13) is
not true. Despite all this, we continue to use this equation because it is both simple
and it is very close to being correct. Does this approximation come up with a close
answer every time we use it? I have had this question in mind for a long time, but in
all the books I have read on this topic, I have never found a satisfying answer. But
I finally realized that there is a universal inequality in the financial world, based on
(8), which I will discuss in the next section.
Reasons Why Business Needs Mathematics
I would like to refer back to the Eq. (8):
1 > DF0.5y > DF1y > DF1.5y > · · · > DFny > · · · > 0.
If you take this sequence and add every second term starting from DF0.5y and
compare it to the result of doing the same process starting from DF1y , you get
DF1y + DF2y + · · · + DFny < DF0.5y + DF1.5y + · · · + DF(n−0.5)y. (21)
By adding the left-hand side of (21) to both sides of (21), you get
2(DF1y + DF2y + · · · + DFny ) < DF0.5y + DF1y + DF1.5y + · · · + DF(n−0.5)y + DFny.
Next I divide both sides by 2(1 − DFny)(> 0) to yield
DF1y + DF2y + · · · + DFny DF0.5y + DF1y + DF1.5y + · · · + DF(n−0.5)y + DFny

< .
1 − DFny 2(1 − DFny)
Using (15) and (17), you get

1 1
< .
pn cn
Since cn > 0,pn > 0, you get
cn < p n . (22)
Similarly, in (8), if you start by adding every second term starting at the left, and
compare this to the result of doing the same process starting from DF0.5y , you get
DF0.5y + DF1.5y + DF2.5y + · · · + DF(n−0.5)y < 1 + DF1y + DF2y + · · · + DF(n−1)y.

(23)
As before, you can deduce that

cn cn
1− < .
2 pn
In the real world, cn is small enough to ensure that 1 − c2n > 0 is true, and hence we
can conclude
cn
pn < . (24)
cn
1−
2
Therefore, by combining (22) and (24), we get
cn
cn < p n < . (25)
cn
1−
2
If we take the true value of pn and compare it to the market convention in (13), the
difference of the two can be denoted as εn and satisfies:
c2n
p n = cn + + εn . (26)
4
Substituting this expression in (25), we get
⎞ ⎛
cn
⎜ +
2⎟
c2n 1 c2n
− < εn < ⎜⎝
⎟. (27)
4 4 cn ⎠
1−
2
The right-hand side of this equation is
14 Y. Yasuda
⎛ ⎞
cn
c2n ⎜ 1 + ⎟ c2 2

⎜ 2 ⎟ = n 1 + c + cn + · · · . (28)
4 ⎝ cn ⎠
n
4 2
1−
2
As |cn | 1, cn to the power of 3 or even higher is much smaller than c2n , we see
c2 c2
that (28) is almost equal to 4n . Therefore, we see that |εn |is of an order at most 4n .
By this calculation, I was able to understand that the market convention (13) is
very accurate since the inequality of (25) holds. However, this equation has no
relevance in a financial transaction, so it is difficult to discover this fact in normal
daily life. But it is still very important to know that (13) guarantees a certain degree
of accuracy. Therefore mathematics is a tool of thought essential to doing business
in finance.
Last Words
Equation (13) has been around since before computers were widely used in the
financial industry, and it is still being used today. Nowadays, when more complex
equations are used in the industry, the fact that this concept has survived is simply
remarkable. Behind all this, there exists this universal inequality (8),
1 > DF0.5y > DF1y > DF1.5y > · · · > DFny > · · · > 0
cn
and as a result, this inequality (25) cn < pn < will always be true. Without
cn
1−
2
mathematics, I would not have been able to understand all of this.
Acknowledgements The author would like to express special thanks to Mr. Hiraku Kusaka, who
is a former colleague at BNP Paribas and is currently working for Nomura Securities Co. Ltd.,
for his full support in preparation of the original manuscript in Japanese and its translation into
English.
Company Profile of BNP Paribas
BNP Paribas is a world-leading banking group headquartered in Paris. As of July

2010, it has around 200,000 employees in more than 80 countries/regions. The
group occupies leading positions in three significant fields of activity: Corporate
and Investment Banking, Asset Management & Services, and Retail Banking.
BNP Paribas also has a strong presence in Japan. Backed by the group’s great
lineups of banking and financial specialists, the Japanese branch provides a wide
range of products and services, mainly in the fields of Corporate and Investment
Banking.
A Brief Biography of Yusuke Yasuda, BNP Paribas/Tokyo
Yusuke Yasuda, BNP Paribas’ former Chief Country Representative for Japan, was
born in Tokyo in 1947. He graduated from the University of Tokyo with B.D.
in Economics in 1970 and earned MBA from the Harvard Business School in 1978.
After working for Japan Airlines, he became Senior Representative of Compagnie
Financière de Paribas’ Japanese office in 1985. In 1989, he was assigned as Chief
Country Representative for Japan and General Manager of BNP Paribas Securities’
Tokyo branch.
In 2005, Yasuda was awarded the Chevalier de la Légion d’Honneur by the
French government. He retired from the company in January 2010.
Takeda Pharmaceutical Co., Ltd.
Yasuchika Hasegawa
President and CEO
Takeda Pharmaceutical Company
Are Numbers Useful?

Acknowledging the Contribution of Mathematical Modeling
to Corporate Management
The Realities of Corporate Management
All companies face a wide spectrum of risks and uncertainties that may impact their
ability to achieve strategic and operational goals. Each and every day, executive
managers confront a myriad of issues that must be analyzed and resolved. Such
analysis is critical in making informed decisions about the future. No matter how
talented or experienced, no company president should make such decisions in
isolation. The organization of a business thus allows a company to allocate roles and
responsibilities and to delegate authority. Through this system and the integration of
specialist knowledge and skills, a company is best able to handle the issues it faces.
Experience is another key factor that can provide valuable insights on future
decisions. Such experience allows businesses to classify problems by type in support
of decision making and to document problems for convenient reference should they
re-emerge. Assessing the policies and outcomes of both successful and unsuccessful
decisions assists future decision making. We at Takeda practice such a policy as part
of our ongoing pursuit of operational excellence. Even so, corporate managers are
called on to deal with many new challenges on a routine basis.
While the above examines one aspect of corporate management, it would be
wrong to suggest that executives and managers simply embrace the qualitative intu-
ition derived from experience in reaching key decisions. It is here that the enormous
18 Y. Hasegawa
contributions are made by quantitative factors. Here, mathematical modeling, which

is the subject of this discussion, comes into play. Broadly speaking, mathematics
contributes to effective decision making in two ways: through statistical analysis of
historic results and forecasting as well as through the process of logical reasoning.
The following sections offer a personal take on the ways in which mathematics
informs the decision-making process and the contributions it makes to corporate
management. I look forward to hearing from and working with professional
data analysts as we work to enhance our business operations and improve our
contribution to society.
The Mathematical Methods Employed by Takeda
Takeda is a pharmaceutical company. Like most companies, it has a number of

administrative departments which are responsible for functions including corporate
strategy and planning, finance and accounting, human resources, and legal affairs. In
our capacity as a leading global pharmaceutical manufacturer, we have also estab-
lished manufacturing and marketing divisions across North America, Europe, and
Asia. The well-developed research and development capabilities and infrastructure
are characteristics of pharmaceutical manufacturers and serve to set us apart from
other industries. Takeda’s research laboratories and clinical development sites are
located in Japan, North America, Europe, and Asia.
Which of these departments performs mathematical operations? I would argue
that they all reap the benefits afforded by quantitative data and mathematical mod-
eling, although only a handful of the departments have substantial direct exposure
to such data. Nonetheless, the benefits and influence afforded by mathematics are
far-reaching in their effects. Let’s take a brief look at some concrete examples of
mathematics at work in the Takeda organization.
Corporate Strategy and Planning
A key department in the management of our operations, the Corporate Strategy and
Planning Department uses sophisticated quantitative modeling. Long-range decision
making is essential to the survival and growth of every corporation. These decisions
must be based on both qualitative and quantitative analysis. In isolation, qualitative
analysis is short on clarity and specifics and has a tendency to be based on hunches
and intuition. It is here that numerical data come into play, primarily in the form of
a technique known as discounted cash flow (DCF) analysis.
DCF analysis is a valuation method that generates future free cash flow pro-
jections, which are then discounted at a fixed rate to arrive at a present value. This
technique is used by companies in all industries, and can be elaborated by deducting
initial investment costs and multiplying the probability of success rate at each stage
of the business. The DCF technique makes it possible to ascertain and compare
Takeda Pharmaceutical Co., Ltd. 19
the value of respective businesses with very different revenue and cost profiles over
time. Added to which, we have recently begun making use of so-called Monte Carlo
simulations to estimate the probabilities involved in bringing a successful new drug
to market.
Sales and Marketing
As might be expected, our Sales and Marketing Department handles many fig-
ures, including sales targets, and these data are generated through mathematical
modeling. In drafting sales and marketing plans, binomial distribution and other
mathematical techniques are used to project sales turnover and market size. Math-
ematical methods, including statistical and stochastic techniques, are particularly
valuable when there is a need to perform a quantitative analysis of market needs.
Research and Development
Research and development is of the utmost importance for pharmaceutical compa-

nies like Takeda and, here too, mathematics has a role to play.
There is a piece of laboratory equipment that is unique in that it bears the
name of a mathematician. Fourier-transform near-infrared (FT-NIR) spectroscopy
is commonly used in laboratories. Does this mean that the researchers who work in
laboratories equipped with FT-NIR spectrometers are expert mathematicians?
The Fourier transform is a mathematical operation that was created by the French
mathematician and physicist Joseph Fourier as a tool for performing the harmonic
analysis of waveforms when solving thermal equations. Microcomputer-based near-
infrared spectrometers for performing Fourier transforms allow researchers to get
results without requiring the branch of mathematics known as harmonic analysis.
This way, they can concentrate on measuring the chemically synthesized active
pharmaceutical ingredients they are observing. Fourier may not have had this goal
in mind, but his development of the Fourier transformation method of materials
analysis constitutes a significant achievement in that it allows researchers to focus
on the task of research.
I would like to conclude pointing out how mathematics is used in pharmaceutical
development to verify the health effects of the active ingredients—the chemicals—
used in our pharmaceutical products. Here, biostatistics is an essential part of the
clinical trials process and is critical to gaining approval for manufacturing and
distribution licenses in conformity with drug regulations.
As demonstrated, mathematical modeling is used in various areas of the phar-
maceutical industry business. However, as the examples cited show, few of our
departments have any direct dealing with mathematical operations. Rather, they em-
ploy one of two approaches: either making mathematical substitutions for a working
20 Y. Hasegawa
hypothesis or using mathematical reasoning as the basis for decision making. It is

my belief that translating these operational tasks into mathematical operations
and converting the math back into practical business requires a knowledge and
understanding of both mathematics and business.
Mathematical Reasoning in Business
The previous section examined the business applications for a number of mathemat-
ical techniques. My goal in this section is to address mathematics and its role within
the company from my perspective as president of Takeda. This can be expressed by
the following equation:
aX 3 + bY 2 + cZ = 0
Reference was made to this topic at the beginning of this article, but a company
president is required to deal with numerous issues spanning a diverse spectrum of
operations in which he or she may have little or no experience. Further, the leader
must be able to reach good, timely business decisions that affect the interest of both
employees and stakeholders and the general public. There is no one “solution” to
the above equation. The right “solution” is wholly dependent on the specifics of any
given set of circumstances.
In managing day-to-day operations, there are often times when I get the feeling
that I am subconsciously engaged in solving a system of equations. These can be
multivariate equations that have a significantly higher number of variable parameters
than those given in the above equation, equations with many thousands of constants
and variables. There are times when it is literally impossible to pin all the variables
down.
Often I have attempted to use my judgment to convert a number of the many
variables into named constants. It is a process that involves bringing the various
variables within a defined area in a bid to stabilize the variable parameters and is
based on suggestions from the numerous individuals who offer their experience. It is
also a process of logical thinking that is essential to reaching quick, timely decisions.
In this sense, the mathematics performed by a company president involves taking
certain variables and having the readiness and resolve needed to transform them
into constants.
While my contact with mathematics may pale in comparison with that of those
who devote their professional careers to crunching numbers, the fact that I am
dealing with forms of mathematics might be interesting to some. Already, significant
progress has been made with Enterprise Resource Planning (ERP) and other systems
that support corporate management, and I am sure I am not alone in anticipating
further support for business from the field of mathematics.
Takeda Pharmaceutical Co., Ltd. 21
Last But Not Least
In closing, I would like to touch upon the subject of the liberal arts. The importance
of a liberal arts/humanities education to corporate leaders and managers is well
established. This is grounded in the fact that in classical antiquity the phrase “liberal
arts” originally denoted those subjects of study that were “intended to liberate the
soul.” Liberal arts refers to the Trivium (the three core subjects taught first) of
grammar, rhetoric, and logic (dialectics) and the Quadrivium (the four subjects
taught following the preparatory work of the Trivium) of arithmetic, geometry,
astronomy, and music, with the three subjects of logic, arithmetic, and geometry
comprising the all-important mathematical elements of the liberal arts.
In reviewing the topic of mathematics in business, I have reaffirmed the
importance to business leaders, who are required to make and be accountable for
sound, independent decisions, of developing reasoning capabilities in mathematical
domains. Going forward as well, I look forward to continuing to forge links with
those who make mathematics their profession.
Company Profile of Takeda Pharmaceutical Company Limited
Founded in 1781, Takeda Pharmaceutical Company Limited is a leading Japanese

pharmaceutical company that in 2011 celebrated 230 years of business. In seeking
to become one of the world’s leading research-based pharmaceutical companies,
Takeda has established a global presence that is founded on an integrated system
of management covering all aspects of the business from research and product
development to the delivery of essential information to patients and their physicians
while striving to realize its vision of contributing to better health for patients
worldwide through leading innovation in medicine.
The company continues to channel its resources into three core therapeutic
areas: lifestyle-related diseases (obesity, diabetes, and arteriosclerosis), oncology,
and central nervous system diseases (depression, schizophrenia, and Alzheimer’s
disease).
A Brief Biography of Yasuchika Hasegawa, Takeda Pharmaceutical

Co., Ltd.
Born in 1946 and a native of Yamaguchi Prefecture, Mr. Hasegawa joined Takeda
Pharmaceutical Company Limited in 1970 having graduated from the Waseda
University School of Political Science and Economics. He has been a Member of the
Board of Directors at Takeda since 1999 following more than a decade of executive
positions at overseas subsidiaries.
Mr. Hasegawa was appointed President and CEO in 2003 having served as
Managing Director of the Corporate Planning Department as well as the Corporate
22 Y. Hasegawa
Strategy & Planning Department. He also serves as Chairman of the Keizai

Doyukai (Japan Association of Corporate Executives), where he was responsible for
publishing the 16th Corporate White Paper on “Creating a New Style of Japanese
Corporate Management” in 2009. Mr. Hasegawa was also Chairman of the Japan
Pharmaceutical Manufacturers Association (JPMA) in 2010.
Nippon Telegraph and Telephone Corporation
(NTT)
Norio Wada
Director and Chairman (2007–2012)
NTT Group
Mathematics Drives the Economy
Introduction
It was not until after I began working that I came to understand the significance of
learning mathematics—a subject in which I was out of my depth and on which I was
destined to spend a huge amount of time. A handful of people will delve into this
subject as mathematicians, but what does mathematics mean to the average person?
Think of the following question as an example. There are one red crayon, one
red pencil, and one blue felt-tipped pen in front of you, and you are asked, “How
many?” How do you answer? It is said that, when small children are asked this
question, many are stuck for an answer. However, if they are somewhat older, it
is said that they can answer “Three” based on the reasoning that all the objects
are writing materials, or they focus on color and reply “two red ones and one blue
one.” Answering this question involves two important thinking processes. The first
is to analyze the objects placed in front of you from a variety of perspectives,
such as shape, color, and purpose. The second is to discover what these objects
have in common. By learning mathematics using numbers and graphics, we have
acquired these thinking processes, i.e., the ability to look at objects from a variety
of perspectives, and ultimately discover a truth that suits the intended purpose and
express it in a simple way.
Once we graduate from school, we are quite often confronted with problems
that we have not studied in school, and become highly upset due to the fact that
we don’t have a clue as to what to do. My experiences show that the mathematical
24 N. Wada
thinking mentioned above is very helpful in solving problems in such situations. In

other words, mathematics is a tool that enables us to broaden our ability to deal with
problems that confront us, and makes our conclusions highly convincing. It is often
said that education is a kind of “multiplication.” Even when both the teacher and
the learner have high potential, if the motivation of either party is zero, the result is
zero, i.e., no education is accomplished. Similarly, this simple rule of multiplication
fully expresses the key to personnel management or personnel allocation. I believe
that, in this way, mathematics provides people with a means to build their own
philosophy. In this respect, I’d like to emphasize the importance of mathematics
education in laying a foundation for teaching people to look at and think about
things from perspectives not confined to any existing framework.
Mathematics Supports the Safety and Security of the Services

Provided by NTT
It is not an exaggeration to say that it is mathematics that ensures that the telecom-
munications provided by NTT are safe and secure. “Safety” in telecommunications
means that you know for sure that you can be connected to the person you want
to contact at any time you want. To make this possible, it is necessary to forecast
the volume of communications traffic, to invest in the required communications
facilities, and to provide the ability to set up alternative routes automatically when
some facilities fail. These activities are supported by a fundamental research area
called traffic theory. Let me introduce a formula that plays a central role in this
theory.
The probability at which the originator of a call is not connected to the called
person is referred to as the call loss rate. NTT has aimed to keep the call loss rate
below 10% in building its telephone network. One mathematical formula important
in the planning and deployment of telephone network facilities is called Erlang B.
It enables us to calculate how many circuits are needed to keep the call loss rate
below a certain level when serving a specified number of users.
aS
B= S!
S
an
1+ ∑
n=1 n!
B: Blocking Probability
a: Traffic (in erlang)
S: Number of lines
NTT 25
This formula was proposed as far back as 100 years ago by Agner Krarup Erlang,
a Danish mathematician. Even today in the 21st century, the Erlang B formula
remains highly relevant and is playing an important role in telecommunications.
This is due to the fact that, simple as it is, this formula very precisely models the
loss rate of telephone calls.
The Erlang B formula has been supporting NTT’s mission of providing network
services as a social infrastructure, and, moreover, NTT has researchers with suffi-
cient mathematical capability to use the Erlang B formula skillfully and are highly
motivated to apply it to the company’s huge telephone network. The “collaboration”
between mathematics and skilled researchers is the foundation on which NTT stands
today. This is also a prime example of how mathematics has been turned into a
practical tool.
Today, the conventional telephone is well past its heyday, and the Internet has
taken over. However, today’s Internet is mostly best-effort-based, and so it is taken
for granted that communication becomes slow or even unavailable when there is
traffic congestion. What is desired from now on is the construction of the NGN
(Next Generation Network), which adds high reliability and a quality of service
(QoS) guarantee to IP communication used in the Internet. In this new environment,
there is a renewed recognition of the importance of the Erlang B formula. Today,
a small number of users who are highly knowledgeable in ICT (information and
communications technology) can use most of the available network bandwidth.
If we can create an environment that retains the beneficial features of the current
Internet, namely, imposing few constraints and allowing use for any purpose, but
can also control traffic in such a way that anyone, irrespective of his or her level
of ICT literacy, can impartially receive network services whenever he or she wants,
then we can greatly contribute to building a better society.
Mathematics That Gives Impetus to Innovation
Mathematical thinking plays an important role in creating services that open new
horizons. It is hoped that ICT will be harnessed to create new innovative services
and build new social infrastructures, with a view to resolving the mountain of
social issues that confront Japan and the world. In this section, let us consider
the importance of mathematical thinking in launching a new business. Starting
a new business requires a variety of capabilities, such as the ability to sniff out
business seeds, willingness to meet the challenge of exploring untrodden paths
as a trailblazer, and the ability to discover some rules in the business you are
exploring (even though you may be groping for them in a kind of fog) and to
develop best practices from them. As I wondered how and what kind of people
acquire these capabilities, I came to realize that there is a clue in geometry. From
two-dimensional figures that you draw on paper with a pair of compasses and rulers,
three-dimensional figures, such as a sphere, to complex multi-dimensional figures
that you can draw only in the imagination, geometry has expressed its intrinsic rigor,
26 N. Wada
unpredictability, and beauty. I believe that the ability to create businesses arises from
geometric thinking. When you look closely at the individual tasks you handle every
day, you’ll find that they exist more or less independently of each other, and it is
often assumed that the ability to judge and solve problems on the spot is important.
This is particularly true at the initial stage of launching a business. However,
this approach is not conducive to creating services or businesses characterized by
consistently high quality and that make a meaningful social contribution. Rather,
it is advisable to take a renewed look at the flows and relationships among people
and data, and create a geometric figure from them. This will enable you to see what
has previously been hidden from you, and see an efficient and reasonable workflow
emerge as a beautiful figure. In other words, I believe that, rather than considering
tasks as a series of steps, imagining a geometric figure that represents tasks as parts
of an ecosystem is essential to solving various issues that arise in our globalized
complex society.
As a specific example, let me introduce a current research project.
One of the core technologies that support service creation is video coding.
MPEG2, which is now widely used in digital television and DVD, was standardized
by ISO/IEC in 1994. The activities of Dr. Hiroshi Yasuda (then an NTT employee),
who worked from a global perspective, played a large role in this standardization.
Today, nearly 20 years on, as data traffic increases day by day, research on
video coding with a higher level of compression is being undertaken globally, with
NTT playing a significant part in it. The current video coding is based on two
technologies. Simply put, one technology reduces the amount of data by regarding
changes in shading within each video frame as waveforms, and by expressing the
frame as a set of superimposed simple waveforms using discrete cosine transform,
and the other technology decreases the volume of data by extracting only changes
from the previous frame in encoding a frame. Conventional video compression
technology has been developed through trial and error by many pioneers in the
field who relied on physical phenomena that people can recognize, such as color
and motion. In contrast, NTT is now doing research on the ultimate compression
technology. This is not an extension of any conventional method but is entirely
different. It uses what is called genetic programming, which can, through iteration of
computation, evolve a video coding function into one providing higher compression.
In other words, unlike conventional technology that provides average compression
performance for any type of video, a technology that can be compared to ready-
made clothes, the new technology aims to provide ideal compression performance
for each type of video, which can be likened to tailor-made clothes.
When seen from a different angle, this research is expected to bring benefits
beyond just high compression of video. Since the coding function derived by
this technology is a mathematical formula that represents a recognizable object,
it has the potential to shed light on the mechanisms by which humans recognize
color, shape, or even an object. A study that was begun to remove reliance on
human perceptual recognition may turn out to help illuminate the mechanisms of
recognition in humans. I have a strong sense of how mysterious this ecosystem-like
NTT 27
or geometric approach is. It is intriguing to wonder what this research, albeit still
in its infancy, which was initiated to compress images or video, will bring us in the
future.
From Mathematics to Corporate Management
In 2004 when I was its president and CEO, NTT announced its medium-term
management strategy. The main thrusts of the strategy were the promotion of
optical access, construction of the NGN, and the provision of new services that
exploit these. At that time, despite its eventual success, aggressive investment in the
construction of optical access networks was derided as crazy by our competitors and
investors around the world. However, I had a firm belief that progress in broadband
communication made possible by the penetration of optical access would usher in a
paradigm shift, the impact of which would be as profound as that of the industrial
revolution. However, when I needed to make a critical decision as the leader of a
huge enterprise, sometimes my determination almost crumbled under the pressure.
Yoshida Shoin led a private school, Shoka Sonjuku, which exerted a great
influence on the Meiji Restoration. Two days before his execution, he began to write
Ryukonroku, which is considered to be his last will and testament addressed to his
students. In this essay he said, “Form a resolution and make it the foundation for
everything.” I assume he meant that everything hangs on how you ask yourself what
you should do and how you form a resolution based on that. This is the most difficult
thing to do today when things change dramatically in a discontinuous manner, but as
such, I believe, it is all the more important for leaders to show this kind of resolve.
Time and again I was encouraged by Shoin’s words, which were backed by
his firm conviction. Today, 7 years after I made that decision, seeing broadband
access widespread and the Internet flourishing, I feel relieved to know that my
determination and actions in those days were not misdirected.
The first part of this chapter described how mathematics and mathematical
thinking play an important role in various activities of NTT and are critical to
ensuring the sustained development of our business. In the concluding part, I
explained how it is necessary in corporate management to forge a future vision
and have the strong resolve to pursue it. The message I wish to convey here is that,
regardless of whether your business is big or small or which industry you are in, it is
only when the above-mentioned two elements—mathematical thinking, and vision
linked to determination—work together like two wheels of a cart that they provide
a great driving power to open new horizons.
Company Profile of Nippon Telegraph and Telephone Corporation (NTT)
NTT is Japan’s largest information and communication company. It has around

195,000 employees in 625 locations worldwide.
28 N. Wada
NTT had been publically managed as Nippon Telegraph and Telephone Public
Corporation until it was privatized in 1985 under the liberalization policy for
telecommunications. In 1999, it created a holding company and shifted to group
management. The group has a significant presence in various fields of business, in-
cluding Regional Communication, Long-distance & International Communication,
Mobile Communication and System Integration.
In the fourth quarter of FY2009, NTT made consolidated net sales of 10.1 trillion
yen and a group operating profit of 1.1 trillion yen.
A Brief Biography of Norio Wada, NTT
Norio Wada was born in 1940. Following graduation from Kyoto University with a
B.A. in Economics, he joined NTT in 1964. After he became a Senior Vice President
in 1992, he held various important posts, including the manager of the Tohoku
Branch and a group Executive Vice President. He became a Senior Executive Vice
President in 1999 and President and CEO in 2002.
Under Wada’s leadership, the NTT Group announced in 2004 the Medium-Term
Management Strategy, which set a company goal to establish the Next-Generation
Network (NGN) based on optical access networks and full-IP telephony. It con-
tributed to the spread of broadband services in Japan.
He was the NTT Group’s Director and Chairman from 2007 to 2012. He had also
experienced being a Vice Chairman of the Japan Business Federation (Keidanren)
for 4 years beginning in 2003. In that position he was in charge of policies for
housing problems and the aging society. Because he has held various important
posts in employers’ associations, he is celebrated in Japan’s business communities.
Nomura Holdings, Inc.
Kenichi Watanabe
Group CEO (2008–2012)
Nomura Holdings
The Role of Mathematics in Finance: Applied Mathematics

and Risk
As far back as the Edo period (1603–1868), rice futures were already being traded at
the Osaka Dojima Rice Exchange. Japan in those days was clearly highly innovative,
given that it was more than 100 years before Europe began conducting similar
financial transactions at the end of the 19th century. However, it was North America
and Europe that led the way in derivatives, starting in the 1970s. Through the
remarkable developments in the field of financial engineering, complex derivatives
and other financial instruments have only become widely used in the last 20 or 30
years. Today, with its extensive use of mathematics, the discipline is probably better
described as “mathematical finance”. Indeed, without mathematical application
finance would not be as sophisticated or globalized as it is now (although the
development of rules, legislations and other infrastructure has also been vital).
Mathematics is essential for analysing financial instruments related to deriva-
tives, a prime example being the application of probability theory. “Ito calculus,” the
stochastic integral and stochastic differential developed by Dr. Kiyoshi Ito, stands
up alongside Newtonian calculus in terms of its importance to probability theory.
“Ito’s lemma” (Ito’s formula) and the Ito representation theorem in particular are
extremely useful in the area of finance, and it would be no exaggeration to say that
researchers and professionals worldwide use these formulae on a daily basis.1
1 The Nomura Group has endowed chairs and research centers at universities in Japan and abroad,
contributing to society. In 2007, to commemorate the 2006 Carl Friedrich Gauss Prize awarded to
30 K. Watanabe
Derivatives function as an effective risk transfer tool because they bring about
the composition and decomposition of future cash flows linked to a reference index
with a specific risk. Derivative transactions between parties wanting to hedge risks
such as equity, interest rate, and foreign exchange rate and parties wanting to take
on that exposure make it possible to effectively transfer the risks. Derivatives can
also be used to vary leverage,2 enabling a financial institution to take on exposure
many times greater than the actual amount of equity capital.
However, as seen with the collapse of Long-Term Capital Management back
in 1998 and more recently at major investment banks where leverage exceeded
30 times, a financial crisis induced a credit crunch, leaving these institutions with
insufficient capital buffers. This highlights that a key management issue for financial
institutions is to ensure the efficient allocation of risk capital. This requires precisely
calculating the potential profit, loss, and degree of risk associated with liquidity
procurement, asset valuations (mark to market), as well as the necessary amount of
equity capital, conducting a factor analysis, and calculating the investment efficiency
(expected excess return per unit of risk). In recent years, the need to upgrade risk
management has become increasingly pressing, but this is not just a simple matter
of calculating the value at risk (VaR),3 the Greeks,4 or credit exposure.5
Returning to derivatives, various types of exotic derivatives (derivatives with
complex path dependencies) have been developed since the end of the last century,
where the application of mathematics has enabled numeric calculations for pricing.
Even for standard (plain vanilla) products like European options, the market has
transformed into one, whereby different strike prices and maturities change the
implied volatility (a phenomenon known as “volatility skew” or “volatility smile”),
spurring competition to develop pricing models with a greater level of sophistication
than the Black–Scholes model. New credit derivatives, such as credit default swaps
(CDSs) and collateralized debt obligations (CDOs), that use credit as the underlying
asset (reference index) have emerged, causing a sudden burgeoning in market
size, but the subprime mortgage meltdown has also made apparent a number of
Dr. Kiyoshi Ito, Nomura Holdings endowed a research division known by the acronym KIMF at
Kyoto University. The division, which operated from 2007 to 2010, comprised two subdivisions
belonging to the university’s Research Institute for Mathematical Sciences and Institute of
Economic Research, respectively. Offshore, the company has established the Nomura Centre for
Mathematical Finance within the Mathematical Institute at the University of Oxford.
2 Leverage is the reciprocal of the equity capital ratio, using the lever principle to increase
investment efficiency. Debt can be used to procure funds to create asset holdings many times greater
than equity capital, boosting the expected return on equity (if the credit risk is excluded).
3 Value at Risk is the maximum expected loss over a certain period of time within a certain
confidence interval. VaR is used in practice to determine standards for the necessary amount of
equity capital.
4 “The Greeks” are the quantities representing the sensitivities (partial differential) of the price of
financial instruments to a change in the underlying asset prices or parameters on which the value
of an instrument is dependent.
5 The amount of exposure to the credit risk of borrowers, counterparties, etc.
Nomura Holdings, Inc. 31
embedded problems with securitized products like CDOs. One example is the
normal Copula model widely used as a de facto standard for risk assessment
(although the assumptions are slightly unrealistic).
The recent financial crisis, spurred by a credit crunch, has served as a reminder
of something that has in fact long been known, namely, that risk diversification
based on the “Law of large numbers” does not always work. Obviously, the crisis
was less about the pros and cons of Copula models, and more to do with the
ways in which securitized products were structured and distributed, spawning
conflicts of interest and fraud-like moral hazards. Market participants as a whole
(as well as policymakers, regulatory authorities, rating agencies, and home loan
borrowers) were also implicated for professing themselves to be diversifying risk,
while ignoring the possibility of credit risk arising, based on the assumption that
real estate prices would continue to increase.
Mathematics leads to simple and beautiful conclusions by ignoring and idealizing
various factors in order to abstract. However, reality is both more complex and
discrete, and in many cases mathematical conclusions are only approximately valid.
There is also the risk that the assumptions used for valuation models differ from the
market (model risk), as well as issues with the reliability of the parameters used in
the model (parameter risk).
Naturally, mathematics is only one of many tools used in finance. Simply
applying mathematics without analyzing economic fundamentals can be not only
futile, but even dangerous. For example, quantitative analysts, or “quants”, tend
to concentrate on formulaic mathematical analyses, putting less importance on the
underlying assets, financial markets, or the real economy. Investors only look at
actual probability (physical measure), but derivative traders and quants analysts
only look at the risk-neutral probability (risk-neutral measure). It could be confused
that in principle mark to market should be done under the risk neutral measure Q,
and risk sensitivities should be measured under the physical measure P. (Even this
dualistic approach is of course overly simplistic.)
Neoclassical economists believe that as the formation of future rational expec-
tations determines the market price at the current point in time, all information is
reflected in the market price. Financial mathematics in relation to derivatives has a
similar conception, where pricing is done under a valuation measure based on no
arbitrage opportunity in the market (risk neutral measure),6 but that is not sufficient
when risk enters the picture. In fact, underlying asset price movements can only
be observed retroactively under physical measure P. In measuring risk, distribution
under physical measure becomes critical.
When it comes down to it, the definition of risk is not obvious, and how to
estimate risk is actually a major challenge. To simulate possible future scenarios
requires many sample paths, but the actual price path is no more than the realization
of one of the envisaged sample paths. Even if past investment management
performance has been strong, it is difficult to determine whether that was through
6 “No arbitrage” means that reliable positive excess returns cannot be obtained without risk.
32 K. Watanabe
skill or luck. This would imply that understanding risk is far from straightforward.
As noted in Nassim Nicholas Taleb’s “The Black Swan: The Impact of the Highly
Improbable”, it is difficult to envisage risks that have never been seen before, but
which could potentially occur.
The recent financial crisis has highlighted exposure towards the risk of liquidity
drying up, whereby a decent asset ceases to sell at an acceptable price on the market,
and the correlation risk, which tends to grow contagiously. Risk factors abstracted
in normal valuation models emerged as much greater than estimated, requiring a
rethink of risk management. It can be said that over the last 10 years, as financial
markets worldwide have all been heavily biased towards derivatives, this has meant,
failure towards observing proper risk management.
In recent years, neoclassical economics has fallen slightly out of favor, shoul-
dered aside by new fields such as “Econophysics” and “Behavioral Finance”.
“Econophysics” studies statistical estimation using actuarial science, extreme value
theory or statistical physics, and consequently does not rely on an idealized
stochastic process of the underlying asset. Experience bears out that the return (the
rate of return) on many financial assets does not follow a normal distribution but
rather a fat-tailed one,7 which exhibits extremely higher values at its tail than the
normal distribution does. Not only the diffusion process seen in Brownian motion,
but also, in the short term, the risk of asset price jumps (defaults, booms and
busts) must be taken into consideration. In the case of non-tradable assets, such
as real physical investments, catastrophe bonds, insurance, or illiquid assets, risk
management needs to be even more conservative.
7 However, this fact was known before econophysics, and the hypothesis that stable distribution and
some law have a dominant effect, with some parameter such as a physical constant existing, seems
a little thinly backed.
Nomura Holdings, Inc. 33
Behavioral finance, an offshoot of behavioral psychology, argues that economic

rationality in the sense of maximizing the utility function (within the framework
of neoclassical economics) is not a quality inherent in human behavior, and that
people make decisions instead on the basis of intuition and emotion. In fact,
people do tend to reason in ways inconsistent with the transitive law. However,
because behavioral finance is close to phenomenology and is not designed to explain
economic rationality, its mathematical approach is still in its infancy, this tends to
be employed in marketing rather than in finance. In recent years, various economists
have been exploring behavioral finance, thus mathematics maybe applied further to
try and merge with neoclassical economics in the years ahead.
At this point in time, both these new schools have yet to add more than a little
understanding to the practice of risk management. As with the old VaR, the market
is probably not stable enough for risk values estimated through statistical analysis
using historical data to be entirely valid, so reservations will need to be attached.
The efficient market assumed by neoclassical economics is only an illusion, with
inefficiencies (anomalies) being a fixture in actual markets, where paradigm shifts
and other sudden structural changes occur repeatedly.
In any case, it will be essential that financial institutions upgrade their risk
management. Closer in depth analysis is needed not only in derivatives and other
financial instruments, but also in the market and underlying assets. This will require
both adequate resources and people deeply familiar with both mathematics and
business.
Company Profile of Nomura Holdings, Inc.
In 1925, the securities arm of the Osaka Nomura Bank spun off to establish
Nomura Securities. It reorganized into a holding company and was listed on the
New York Stock Exchange in 2001, while Nomura Securities remained the group’s
core company. Since the acquisition of Lehman Brothers’ Asia-Pacific franchise and
certain operations in Europe and the Middle East, the Nomura group has accelerated
its overseas operations, providing integrated financial services in more than 30
countries or regions. The group not only has achieved a dominant share in Japan
but has solidified its position as a top-ranking investment bank in the world.
A Brief Biography of Kenichi Watanabe, Nomura HD
Kenichi Watanabe is Senior Advisor to the Board of Nomura Holdings and

Chairman of the Board of Trustees of the Nomura Institute of Capital Markets
Research.
Prior to his current position, he was Group CEO of Nomura Holdings, Inc. from
April 1, 2008, to July 2012. Before becoming Group CEO, he headed Nomura’s
domestic retail operations for 4 years from April 2004 to March 2008.
34 K. Watanabe
Mr. Watanabe started his career at Nomura in 1975 in corporate finance and has
held a number of key positions in finance as well as international planning.
He was promoted to the Board of Nomura Securities in June 1998 and has been
responsible for a wide range of roles including finance (CFO), credit evaluation
(CRO), and administrative affairs (CAO), which included general services, compli-
ance, and investor relations.
He is also a member of the Board of Trustees of the Nomura Foundation, which
contributes to advancing the social sciences and the world economy, cultivating the
arts and culture, fostering international talent, encouraging international understand-
ing, and supporting the progress of human society.
Outside the firm, he is an Executive Member of the Policy Board of the
Japan Business Federation (Keidanren), contributing to the growth of the Japanese
business community.
Mr. Watanabe graduated from Kobe University with a B.A. in economics. Away
from the office, he is an avid reader and particularly enjoys Chinese classics.
HORIBA, Ltd.
Atsushi Horiba
Chairman, President and CEO
HORIBA
Mathematics Is the Starting Point of Corporate Culture
HORIBA, Ltd. Started as a Student Venture Company
HORIBA Radio Laboratory was started as a business in October 1945 by Masao

Horiba (now Supreme Council) as the first-ever student venture company in Japan.
Later, in 1953, the company was incorporated as HORIBA, Ltd. The first product of
HORIBA, Ltd. was a pH meter, a device for measuring the acidity or alkalinity of a
solution. Today, the pH meter is one of the company’s major products.
Currently, the HORIBA Group has been developing, producing, and delivering a
variety of equipment, including control units and inspection devices indispensable
to semiconductor production; testing systems and emission measurement devices
essential to the development of engines and automobiles; blood diagnostic devices
used in the medical field; atmospheric measurement instruments that contribute to
environmental protection; and analytical devices used in advanced research aimed
at future industrial development. Resulting sales amount to nearly 110 billion yen
(USD 1.4 billion) per year. Many products contribute to production and development
activities for a wide range of industries, from the development of Formula-1 engines
to the development and production of state-of-the-art nanomaterials.
Based in Kyoto, HORIBA, Ltd. has been in business for 67 years since its
founding, with 36 group companies and nearly 5,000 employees spread across the
world. Kyoto, where our company’s head office is located, has been the center of
Japanese culture ever since a capital was established there 1,200 years ago. A very
remarkable point of Kyoto culture is that it accepts only genuine items. Although
36 A. Horiba
an advanced information network such as the Internet brings us diverse information

instantaneously from all over the world, people in Kyoto ascertain the core value
of things. In Kyoto, it is considered important to distinguish the true nature of
things by sometimes simplifying or changing the way of looking at them. Also,
Kyoto’s character is such that assessment of value is seen not in terms of the size
of something but in terms of its content. Such Kyoto culture, in which identifying
the true nature of things is valued, may have provided the fertile ground for many
achievements by Nobel laureates in physics and chemistry based on mathematical
principles.
Products Are Based on Mathematics
HORIBA Group products use chemical and physical reactions to measure the
parameters of substances in various forms including atmospheric air, the water in
our environment, and solid state matters. Therefore, many personnel ranging from
ordinary employees to executives have scientific backgrounds, which means the
education they received in high school and university included mathematics, with
no exception. The development and design of measuring instruments inevitably
involves a large amount of mathematical content. Some of these instruments work
by using mathematics itself.
The pH meter, our company’s first product, is based on electrochemistry. This
product measures hydrogen potential. An analysis device that is one of our key
products measures gases based on the absorption of infrared light, which is a part
of the electromagnetic spectrum. In recent years, efforts have been made to develop
laser-based gas-measurement technology as well. A laser beam whose wavelength is
adjusted to a target absorption peak is radiated on the basis of quantum mechanics,
and gas concentration is calculated from the extinction intensity of the light beam.
As another example, we have produced spectrometers utilizing Fourier transform
technology, which is based on general mathematical principles. This method, which
is called “Fourier transform spectroscopy,” can measure absorption spectra of
substances with high sensitivity within a short time because it uses light intensity at
high efficiency. Here, owing to the wave characteristics of light, beams interfere
with each other. As shown in Fig. 1, by splitting a light beam from a light
source into two sub-beams and subsequently combining them back into a single
beam with a beam splitter, interfered light can be generated and detected. Such
a device is called a Michelson interferometer. If the mirror on one of the split
paths moves back and forth continuously, the optical path difference between the
two sub-beams changes according to the mirror position. In the case of single-
wavelength light, the intensity profile of the combined beam at a detector is
represented by a cosine wave because of interference. In the case of wide-band light,
which consists of continuously different wavelengths, the light intensity becomes
strongest when the optical path difference is zero and weakens with the path
difference. Figure 2 shows the light-intensity profile caused by interference, referred
HORIBA, Ltd. 37
Fig. 1 Schematic of a M1
Michelson interferometer
S
BS
L1 M2
L2
D
S: Light source, BS: Beam splitter, L1, L2: Lenses,
M1: Fixed mirror, M2: Moving mirror, D: Detector
to as an interferogram. An interferogram is composed of cosine waves with various

frequencies corresponding to wavelengths. The interferogram can be converted into
an intensity profile against wavelength, i.e., spectrum, by frequency analysis using
fast Fourier transform.
We also produce an optical device that has similar effects with Fourier transform.
This optical device, diffraction grating, is manufactured by HORIBA Jobin Yvon,
located in France. Ultrafine grooves are formed in parallel on the surface of the
grating. Light beams diffracted by the grooves cause interference at a specific angle,
thereby becoming stronger. With the use of a lens, the light beams can form an
image at the focal position as a spectrum. This means that the light beam is spatially
decomposed into a spectrum just as in a Fourier transform process.
In these two examples, physical phenomena described by mathematical expres-
sions are utilized in analytical instruments.
Identification of the True Nature of Things
When describing natural phenomena, we human beings have been using mathemat-
ical expressions to simplify the principle, not only in physics and chemistry but also
in biology, thereby expressing phenomena intuitively and in an objective fashion.
Here, “to ascertain the essence of things” and “to identify the true nature of things”
which underlie Kyoto culture seem to coincide amazingly well with methods of
mathematical expression.
38 A. Horiba
Intensity
Wavenumber (cm−1)
Fig. 2 Relationship between an interferogram and a spectrum
Naturally, employees in R&D and engineering have scientific backgrounds, and

so do many employees in sales and administration departments. As our customers
are in the fields of science or engineering, sales departments need staff members
who are experts in these fields. Employees well-grounded in mathematics are
assigned to production and administration departments as well, performing their
jobs from a mathematical point of view, finding the core of problems. Employees
with nonscientific backgrounds also naturally acquire a similar way of thinking.
The technologies that the HORIBA Group possesses are like a mass of accu-
mulated know-how. “Human links = chains” form corporate assets. Ever since our
foundation, we have been expanding the range of measurement technology further
and further. In our company, technological chains filled with ideas are laid out like
a web. Speaking of chains in terms of networks, when looking back over the last 20
years, the listed companies in Kyoto were scarcely affected by the economic bubble.
This is because the Kyoto network, which traverses market segments, has a cultural
climate where only genuine information is communicated. In our company, even if
an outside voice says “such and such business is profitable,” the truthfulness of the
information is confirmed by making full use of our in-house engineering, sales, and
business management networks. Consequently, only the true information survives.
Communications
Among a total of nearly 5,000 employees within the HORIBA Group, about
3,000 speak languages other than Japanese. The number of French employees is
particularly large at 1,000. Ranking next are 700 Americans and 700 Germans,
followed by 200 in China and 150 in the United Kingdom. All the people working at
HORIBA, Ltd. 39
HORIBA group companies including executives are called “HORIBARIANs,” who

are now in 27 countries worldwide.
In many global Japanese enterprises, products are often developed and designed
in Japan and then exported to foreign countries. In the HORIBA Group, there are
R&D and engineering in France, Germany, and the United States. In Austria, South
Korea, and the United Kingdom, modification and customization are carried out
independently so as to optimize the products for local demands. In terms of sales
volumes, the Japanese market accounts for 30% and foreign markets account for
the remaining 70%. How to capture customers in foreign countries is a key point in
maintaining and increasing our competitiveness.
The technical network in our company is formed on the basis of communications.
Accordingly, communication between employees and executives is particularly
important. In Kyoto, a buffet-style birthday party hosted by executives is held every
month after working hours, inviting about 100 employees from Japanese HORIBA
group companies whose birthdays fall in that month. In the Tokyo metropolitan area,
a similar birthday party is held once every 3 months. What is unique about these
parties is that no middle managers are invited. First, I give a talk about recent events
in the company, business results, investor ratios, etc. to the participants ranging from
new to experienced employees. After that, the employees go on to have enjoyable
conversations with executives.
In 1996 and 1997, we acquired two French companies. In 2005, we acquired the
business of a German company. As a result of the acquisitions, foreign employees
account for 60% of all employees. However, the HORIBA Group corporate culture
is still maintained as a whole, which is our great advantage.
In 1991, we established a training facility with a capacity of about 50 people in
Takashima City, Shiga Prefecture. Of course, the main purpose of the facility is to
have seminars and conferences for employees. Also provided is space where, after
the sessions are over, people can enjoy drinking and relaxed conversation around a
fireplace. All the drinks are free and there is no lights-out. Twice a year, dozens
of top managers from Japan and abroad gather at this facility to hold face-to-
face executive meetings. Members from different countries and different business
segments (automotive, semiconductor, medical, environmental and scientific) get
together to hold discussions on business strategies. The greatest benefit from these
occasions is that the participants stay, eat, and drink together, which strengthens
the bonds among HORIBARIANs. In 2009, the training facility was expanded three-
fold in size to provide a capacity for training or meetings of 120 people at any time.
Accommodations at the facility, which used to be in a lodge-style guesthouse, have
been upgraded to a hotel-style resort.
There are three foreign corporate officers in the business management layer
of HORIBA, Ltd. More than 10 years have passed since two of them joined the
HORIBA Group. It is impossible to manage our business by constantly sending
detailed instructions to top executives in companies located across the globe. In
management, how the values of head office are communicated is important. It is
also worth noting that two of our three foreign corporate officers have PhDs: our
French corporate officer has a Doctorate in Science and the other has a Doctorate
40 A. Horiba
in Engineering and has held a V.P. position with a top semiconductor and related
device manufacturer in the United States, which adds a global perspective in our
management team.
Human Assets Are Invisible Properties That Greatly

Support the Group
The HORIBA Group has become a large corporate entity. But we aim to remain a
“great medium-sized enterprise” so that “big company disease” will not develop.
Our company is strictly a vessel for human wisdom, and we place a high value
on this fact. In the HORIBA Group, we call human resources “human assets.”
Engineers tend to portray an inorganic image as their work is based on mathematical
principles. However, when combined with the human nature of being able to
concentrate and focus on one thing, they can work enthusiastically. The skill to
understand and manage the engineers’ minds is a part of the know-how in our
company’s management style.
After studying physics at a Japanese university, I enrolled as a junior in
the Department of Electrical Engineering in the College of Engineering at the
University of California Irvine (UCI). At the time, I had a very hard time coping with
the English language. However, the mathematics I studied at the university in Japan
was helpful. There are no language barriers in mathematics. After graduating from
UCI, I worked as the first HORIBA service engineer stationed in a foreign country
and dealt with a considerable amount of our products that experienced problems.
When I made reports to the head office in Kyoto, the R&D team’s response was,
“We have never had such problems in Japan.” The pain I experienced at that time
could be likened to what must have been experienced by Japanese border soldiers
who were stationed far away from home more than 10,000 years ago. I felt caught
in a crisis, thinking, “Operations that give no thought to the front-line troops will
stagnate.” As a result of this experience, I dared to become “anti-headquarters.” One
result of this has been our spirit of placing importance on the front line that supports
this global company.
In large enterprises, job assignments tend to be divided up according to em-
ployees’ specialties. For example, electricians are assigned to electrical tasks and
physicists are assigned to physics tasks. However, in the HORIBA Group, our
engineers are trained so that every engineer can understand electricity, physics,
chemistry, and software. At present, the number of our product models in headquar-
ters exceeds 1,000. The number increases considerably when the group companies’
products are combined. If 5,000 employees are to handle operations ranging from
development to after-sales service for all our product models, every employee needs
to be multi-functional and become a multi-skilled worker so that they can deal
with any situation. Here also, a mathematical way of thinking seems to be of help.
We have invested significant time and effort in in-house education.
HORIBA, Ltd. 41
Training and education are provided at a variety of levels and fields to develop
human assets that support “technical meshes = chains.” A training program called
“HORIBA College” was also established, consisting of 260 classes spreading across
all departments, strengthening the power of our human assets, and leading to
improved product power and organizational power. Each year 10 or more employees
from group companies in Japan are sent to our overseas group companies for a
1-year internship to develop a global point of view. These employees are selected
not on the basis of linguistic proficiency but on the basis of what they want to do. So
far, this system has given 10% of our employees experience working with different
cultures and ways of thinking in foreign countries. And now, in any department in
our company there is at least one employee who can communicate in English with
experience from living in a foreign country.
In any company, things will not go well with the effort of only a single talented
person. Even if not every employee is a superman, we need to work together to
ensure that all our employees in combination will be like panes of stained glass and
radiate beauty when illuminated by light, making the HORIBA Group as a whole a
super dream team.
The HORIBA Group Is One Company
Sixty-seven years have passed since the HORIBA Radio Laboratory came into being
in 1945. Sales have risen to over 100 billion yen per year, and 60% of the employees
in the HORIBA Group are non-Japanese. There are R&D bases in foreign countries
as well, where managers of local nationalities are in charge of the business. In spite
of the expansion, I am happy to say that the venture spirit cherished since the early
days of our founding has been passed on. This fact is expressed symbolically in
our company motto: “Joy and Fun.” This motto contains the hope that the work
on which we spend the most active period of our lives should be addressed with a
spirit of pride and challenge. In an R&D-oriented enterprise like ours, added value
is created only when employees find their work worthwhile.
The spirit of “Joy and Fun” has penetrated our group companies abroad as well.
They differ not only in ethnic groups and cultures, but also in business fields.
A center-of-gravity force is required to unify these differences and diversities into
a single brand. Unification leads to corporate values and produces a corporate
culture. The products of our company are based on science including mathematical
principles. Our executives and our managers as well as many employees have
scientific backgrounds. Furthermore, about 50 of our employees in Japan and about
100 of our overseas employees have PhDs. The HORIBA Group is basically run on
a global basis by around 10 top executives, including some non-Japanese. Most of
them have engineering experience.
Executives in the HORIBA Group learn their business management strategy by
trial and error. Since they have scientific backgrounds, they can naturally conduct
experiments and go on to implement the outcome on the basis of the resulting
42 A. Horiba
Fig. 3 At an international in-house seminar
data. The process of explaining the natural or social sciences using a mathematical
calculation is called simulation. When an actual simulation is performed, the result
often turns out to be different from reality. Lessons are learned from these failures
and improvements are made. As a result, the simulation approaches reality more and
more closely.
Following the spirit of “Joy and Fun,” Japanese and French R&D personnel
or American and German R&D personnel cooperate internationally to raise the
competitiveness of products bearing the HORIBA brand. Many of the products of
the HORIBA Group hold the largest market shares in the world. These products
cover various fields, including the automotive industry, the medical field, the field
of analytical devices used for environmental measurement and advanced research,
and the semiconductor manufacturing field. The absence of these products would
disrupt automotive development, and the manufacture of semiconductors would be
discontinued all over the world. This reflects our social responsibility and our pride.
HORIBA engineers across the world support our products.
Saying that “the HORIBA Group is one company” means that business is
conducted on the basis of a single set of values and a single corporate cul-
ture. However, the localities of each different region are also respected. Taking
environmental measurements as an example, there are some differences in the
regulations between Japan and Europe. However, engineers and top management
overcome these differences between regions while being connected on the basis of
mathematical principles and one corporate culture. Another important point is that
all entities take pride in their own cultures and learn from one another’s experiences,
forming a strong team. In that sense, it is fortunate that the head office is located in
Kyoto, which has a long history of cultural diversity.
HORIBA, Ltd. 43
Company Profile of Horiba, Ltd.
HORIBA is an analytical device manufacturing company located in Kyoto. It was

founded in 1945 by Masao Horiba (now Supreme Advisor) when he was still a
student of Kyoto University. It was one of the earliest cases of student companies
in Japan. Since the 1990s, it has expanded its scale of operation through M&A
activity to achieve the biggest world market share in emission measurement systems.
The range of business consists of automotive R&D, process and environmental
monitoring, in-vitro medical diagnostics, and semiconductor manufacturing. Its
branches are located worldwide, including in France, Germany, the United States,
the United Kingdom, Austria, and Korea. The HORIBA Group has around 5,000
employees, 55% of them having come from outside Japan. In 2009, it achieved
sales of 104.5 billion yen.
A Brief Biography of Atsushi Horiba, HORIBA
Atsushi Horiba was born in Kyoto in 1948. He graduated from the College of
Engineering, University of California, in 1977. He joined HORIBA in 1972 and
worked as Overseas Executive General Manager, Senior Sales Manager, and Senior
Executive Director before becoming President & CEO in January 1992. Since June
2005, he has been Chairman, President & CEO. He has also held various important
posts outside the company, including President of the Japan Analytical Instruments
Manufacturers’ Association (JAIMA) and a member of the Board of Advisors at
UC Irvine. He received the Officier de l’ordre national du Mérite in 1998 and the
Chevalier de la Légion d’Honneur in 2010 from the French government.
Railway Technical Research Institute (RTRI)
Eisuke Masada
President
Railway Technical Research Institute
Mathematics Supports Development of Railway System

Technology
The Railway Technical Research Institute (RTRI) was inaugurated as a foundation

to succeed the research and development division of the former Japanese National
Railways when it was privatized and separated into seven companies. RTRI is
carrying out comprehensive research and development from the basics to practical
applications in railway technologies and labor science, with the aim of contributing
to the growth of railways. The railway is a complicated mega-system, composed
of various equipment different in character, such as rolling stock, tracks, stations,
signaling and control systems, power supplies, and so on. Its operation and service
are influenced by human factors and must cope with natural disasters such as
earthquakes and high winds. The systems engineering approach is indispensable
to the management and design of such a system. Its basis lies in mathematics. I,
myself, have participated in studies of analysis and design of various systems as
a specialist in systems engineering. Abstract mathematical theory, which I learned
in the general education course in university, has been very useful in solving these
problems. The essence of railways is to operate trains for the transport of people
and goods, after all. For clarification of various phenomena occurring in those areas
and for improvements in performance such as an increase in operational velocity,
it is essential to understand physical motion related to those phenomena intuitively.
It also is based on the knowledge of physical mathematics.
I would like, therefore, to explain the relation between our problems and
mathematics more practically with several typical examples.
46 E. Masada
To Improve Performance of Trains
Objectives such as increasing the speed of trains, better riding comfort for pas-
sengers, and improvement of maintainability of rails are pursued in railway
technologies at all times. The explication of causes of abnormal phenomena such as
derailment is also important. The mathematical analysis of train motion on the track
is the basis for these studies. The train consists of a car body, bogies, wheels, traction
motors, and other parts. It runs following the track profile, and its components are
connected to each other with suspension systems such as springs and dampers to
suppress vibration caused by running. Physically, the running state of a train can
be expressed in analytical dynamics as a multi-body motion, where it is restrained
at the contact point between wheels and rails, and its components are connected to
each other with suspension systems. The nonlinear elements as the friction force
working at the contact point of the wheel and the elastic motion of the car body
itself also need to be considered. Then the motion of the vehicle is given by a group
of high-order nonlinear differential equations. It can be solved only by numerical
analysis. The multi-body dynamics program is usually applied for this purpose.
Final verification is carried out by operational tests utilizing a real train, seeing
whether the analytical results obtained with such a numerical method show practical
phenomena correctly. Because it is difficult to ascertain the mechanical construction
and to adjust characteristics of components in actual operation, the test equipment
called the rolling stock test stand, as shown in Fig. 1, is utilized. Operational tests
are carried out on rolling stock, in which longitudinal motion is restricted and
which operates with wheels driven by rotating discs simulating rails. Their results
are compared with analytical ones obtained with a real-time simulation program,
which runs parallel to and coincident with the operation on the stand. It is possible
to prove the design of rolling stock through such an arrangement before actually
manufacturing and testing it.
In the case of a magnetically levitated transportation system, which is called the
“linear motor car” in Japan and is shown in Fig. 2, the thrust, lift, and guide force
working on the train are determined with the electromagnetic forces generated by
the current flowing in coils equipped on the guide way. Therefore its characteristics
cannot be fully dealt with unless both the dynamics equations mentioned above and
electromagnetic equations are solved simultaneously. It requires a more sophisti-
cated scheme for numerical analysis.
To Cope with Natural Disasters
It is important in terms of the security of public mobility for infrastructure such as

tracks and bridges to withstand natural disasters and to continue stable operation
in our country, which suffers from earthquakes and typhoons. For this purpose, it is
necessary to analyze how a structure vibrates with a seismic wave, how the flow of a
Railway Technical Research Institute (RTRI) 47
Fig. 1 Rolling Stock Test Stand
Fig. 2 Superconducting Maglev under Tests in Yamanashi Test Track

48 E. Masada
Fig. 3 Aseismatic Evaluation and Measures for Existing Railway Facilities considering their
Interaction
swelling river behaves around a bridge pier, and so on. With recent unstable weather
conditions, new tasks are created such as evaluating the influence of strong winds
against a moving train.
Figure 3 shows a model of a situation, where a seismic wave is generated with a
dislocation, propagates in a stratum with a complicated structure, vibrates structures,
and exerts an influence on a running train. If such a phenomenon can be analyzed
numerically with the finite element method, and the state of vibration as well as
its frequency characteristics are clarified, we can consider the proper measures for
secure structures and safe operations. It makes it possible to evaluate measures
against derailment for a train operating relatively near a seismic center.
To Reduce Deterioration of the Environment
An environmental problem caused by the railway system is the audible noise gener-
ated by train operations in areas neighboring the railway line. The aerodynamic
noise from a moving train increases with its velocity. Therefore, it is important
to introduce measures to suppress aerodynamic noise for further increasing the
operational velocity of the high-speed system. The aerodynamic noise is caused
by exfoliating eddies generated at discontinuities in the shape and projections of the
car body as well as by the oscillation of the turbulent boundary layer, which exists
along the whole surface of a train. The sound source can be identified and evaluated
clearly with the results of measurements in a wind tunnel as shown on the left-hand
−104 104
Fig. 4 Wind Tunnel Test and Simulation
side of Fig. 4, or with a running test of a train with a microphone array, in which a
large number of microphones are arranged along the cross section of the car. On this
basis the measures are studied in the design of the shape of a train. Because the level
of audible noise is important in the frequency area, mathematical methods become
important for revising differences in observation windows in the measurement of
each microphone caused by its position and the class of sound source.
Although it is a rather localized matter, the audible noise from the pantograph
(current collector) is remarkable in high-speed operation. Applying porous material
to the surface of the pantograph is being considered in the study of measures to
suppress noise. The right-hand side of Fig. 4 shows the field of air flow obtained in
experiments and simulations in a case where this material is affixed to the surface
of a cylindrical component. It has been proven by solving the field of air flow with
numerical simulation, that only a limited number of stable eddies are generated in
the case where the porous material is affixed. It confirms the improvement of the
performance of the air flow and makes it possible to select the appropriate porosity
of material and its thickness.
To Control Train Operation
The railway system will be reliable for passengers if it is operated following

correctly the scheduling diagram for trains and can recover its scheduled oper-
ation in a short time in the event that abnormal situations such as accidents or
malfunctions occur. It is a problem of programming to draw up the scheduling
diagram, where various conditions are taken into account such as the number
of trains, existence of trains operated with priority, the number of drivers and
the arrangement of their bases, the distance between stations, the composition of
50 E. Masada
Fig. 5 Sketch of Dispatching Information Transfer System
platforms, and the topographical restriction of routes. Because trains of different

lines and companies extend their service routes into one another and the traffic
density is very high especially in the suburbs of metropolises, scheduling becomes
a matter of considerably complicated mathematical programming. The signaling
system, which sets the routes of trains and secures safe operations, should operate
logically interlocking with the scheduling diagram, as well as with the train
information system, which provides information to passengers on departing and
arriving trains. The designing of these systems is also a subject of mathematical
methods.
It is desirable to complete the scheduling diagram, the assignment of drivers for
each train, and so on automatically with computers as countermeasures in cases of
special events and the occurrence of accidents, and to transmit this information to
all posts concerned as occasion demands, because in such cases it is necessary to
cope with the situation in a short time. Figure 5 shows an example of the automatic
process of restoring the time schedule through mathematical programming as the
operational schedule of freight trains is disturbed. If such software is integrated
into the information system that controls train operations, the disturbance in time
schedules can be fixed in a short time.
As shown by the examples above, mathematical methods such as analysis, logical
programming, and mathematical programming are utilized extensively in research
and development in railway technologies. The practical design and evaluation of

the performance of the railway system are carried out mainly with a numerical
calculation scheme. The railway has become more and more complicated as a
system as its network has been extended and its operational velocity has been
increased. Past approaches such as trial manufacture and testing are limited in
effectiveness to comprehend the performance of a system that operates in various
environments and integrates diverse technical fields. It is essential to further apply
mathematical thinking in the abstract and to use simulation analysis as its basis. At
the same time, statistical methods are utilized for understanding various parameter
fluctuations and their time variations in the system for evaluation of characteristics
and for decisions in the management strategy of railways. The importance of
mathematical knowledge increasingly rises as the grounding for research and
development of the railway system.
Company Profile of Railway Technical Research Institute (RTRI)
The Railway Technical Research Institute (RTRI) was established on December

10, 1986, under the authorization of the Minister of Transport (now the Minister
of Land, Infrastructure, Transport and Tourism). This was before the breakup
and privatization of the Japanese National Railways (JNR) took place on April
1, 1987. As soon as Japan Railways (JR) companies, which inherited the assets
and operations of JNR, started their business, RTRI took over its R&D programs
and went into full swing. Its main laboratory is located in Kokubunji, Tokyo.
Its 500 employees devote persistent effort to achieving innovations in railway
technologies, covering both basic and applied research in various fields including
civil, electrical, information, material, environmental, and human engineering, as
well as technologies of rolling stock.
A Brief Biography of Eisuke Masada, RTRI
Eisuke Masada completed his doctoral work in the Graduate School of Mathematics
and Physics, The University of Tokyo, in 1965. As soon as he received his PhD in
Engineering, he started teaching in the university’s Department of Electrical Engi-
neering. He became a professor in the Department of Engineering in 1983. After
his retirement in 1998, he became a professor in the Department of Engineering,
Tokyo University of Science. He has been President of RTRI since 2007 and has
held various posts, such as President of the Institute of Electrical Engineers of
Japan (IEEJ), President of the Electric Technology Research Association (ETRA),
a director of Fuji Electric Holdings and Vice-President of the Japanese Industrial
Standards Committee (JISC). His awards include the IEEJ Meritorious Contribution
Award, the Transport Culture Award (from the Ministry of Transport), and the
EPE-PEMC Council Award.
Japan Oil, Gas and Metals National Corporation
(JOGMEC)
Hirobumi Kawano
President
JOGMEC

Exploitation Industry in Japan
Introduction
In this article, how tasks are theorized, quantized, or transacted in various fields and
areas of petroleum and natural gas exploitation is described. The requirements for
future mathematical development are also explained. Firstly, a brief description of
the features of petroleum and natural gas exploitation is given.
What Are Oil Fields and Natural Gas Fields?
A schematic view showing oil and natural gas fields is given in Fig. 1. The green-
colored area is often described as an oil reservoir, and the figure may be interpreted
as an underground cave filled with crude oil. The reality, however, is that the green-
colored area is a rock layer as seen in Fig. 2. In many cases, the rock that collects
oil or gas is sandstone or limestone (see Fig. 3). There are very small continuous
cavities inside them, and collectively they constitute 20–30% of the volume of all
rocks. Inside the cavity, there may be oil, gas, or water, and together they form an
oil field or a natural gas field.
54 H. Kawano
Fig. 1 Oil–gas field a few thousand meters underground [ref. “Petroleum Development Technol-
ogy Guide”, Japan Petroleum Development Association (in Japanese)]
Fig. 2 Oil–gas reservoir is a rock layer [ref. “Petroleum Development Technology Guide”, Japan
Petroleum Development Association (in Japanese)]
The Origin of Oil Is Not Where It Exists Today
Plankton that accumulates at the bottom of seas and lakes is sedimented deep into
the ground through siltation. During this process, the increasing ground temperature
and pressure with depth causes some portion to change into sources of petroleum
and natural gas (i.e. hydrocarbon compounds). Rock generated by sedimentation
of plankton with soil, typically mudstone, is called source rock. Because of the
Japan Oil, Gas and Metals National Corporation (JOGMEC) 55
Types of reservoir rock Photomicrograph of reservoir rock
limestone
sandstone volcanic rock
sandstone limestone volcanic rock Blue part denotes porosity

for oil and gas
Fig. 3 Types of rock constituting an oil–gas reservoir and its cavity [ref. “Petroleum Development
Technology Guide”, Japan Petroleum Development Association (in Japanese)]
difference in specific gravity from the surrounding water, the source of petroleum
and natural gas thus generated drifts in the rock layer for millions of years until it
reaches its present position, forming an oil or natural gas field.
Method of Oil Field and Natural Gas Field Exploitation
As the specific gravity of petroleum and natural gas is generally lower than that of
the surrounding water, the probability of their sedimentation is higher in convex-
shaped, low-permeability, dense rock with an upper rock layer which acts as a
cap (the dark-colored seal layer in Fig. 4), a rock layer caused by dislocation or
by protrusion of rock salt. Thus, petroleum exploitation begins with a search for
the aforementioned rocks and rock layers deep in the earth. Exploration for rocks
and rock layers capable of petroleum sedimentation at depths of a few thousand
meters underground and of several hundred meters underwater is primarily done
using seismic waves (see Fig. 5). Vibrations generated by dynamite or an air gun
propagate underground and reflect at the interface of rock layers. The underground
structure is then analyzed using data, such as elapsed time, phase, and amplitude,
from the received reflected wave. The principle resembles that of ultrasonography
used in medical practice (see Fig. 6). There are great differences, however, between
searching through subcutaneous fat in a human body, typically on the order of few
centimeters thick under the skin, and through rock(s), which may contain salt water
or oil, typically on the order of few thousand meters under the surface of the earth.
There are many cases in which a slight error may cause considerable changes in the
interpreted profile of the underground structure.
56 H. Kawano
seal
reservoir
rock
Oil field
4 3
2
source rock
1
basement rock
Fig. 4 Drift of petroleum gas inside rock layer [ref. “Petroleum Development Technology Guide”,
Japan Petroleum Development Association (in Japanese)]
Method of Direct Petroleum Exploitation
Even when a rock layer that may be an oil–gas field is found, there is no current
technology to determine whether the rock layer a few thousand meters underground
contains crude oil, natural gas, or just water. The presence of crude oil or natural gas
can be only found by drilling a well into the rock layer. Drilling an exploratory well
costs 100,000,000 to 1,000,000,000+ yen, and may take 3–6 months or more.
Risk Evaluation Accompanying Oil–Gas Field Development
Once crude oil or natural gas has been found by drilling some wells, an estimation
of the size, extent, and quality of the reservoir, calculation of reserves, and
development expenditure is made using the limited amount of information. In oil
exploitation, a huge investment is made initially, and return on the investment
is sought during the production period, which may last for several decades. The
rough shape and extent of the oil reservoir rock is obtained from the seismic
exploration. Porosity and the oil/water ratio is estimated from the electrical and
acoustic properties of rock samples taken from the wells and from the approximately
1 m well-hole made by dropping a logging tool into the well. The average fluidity
of the oil inside the reservoir is estimated from test production and pressure data.
Putting the information together, an oil reservoir model is constructed. The future
satellite
survey
vessel
streamer cable (receiver) receives the reflected sound wave
buoy buoy buoy
streamer cable (receiver) air gun

(hypocenter)
seabed
sound wave
reflects at the
interface of
rock layers
Fig. 5 Principle of seismic exploration (ref. JOGMEC/METI)
Fig. 6 Resemblance to ultrasonography diagnosis (ref. JOGMEC)

58 H. Kawano
Fig. 7 Estimated result of oil-field production by numerical simulation (ref. JOGMEC)
crude oil and natural gas production for various exploitation conditions is estimated
by numerical simulation, and basic data for calculating the economic viability
of oil field development is determined (see Fig. 7). The technology and know-
how manifests itself in performing production estimation with high accuracy from
limited amounts of information.

Exploitation and Development Industry
Features of Information (Data) Treated
Numerical simulations are carried out for various problems, ranging from exploita-
tion to development and production of petroleum and natural gas resources. In
general, handling strata a few thousand meters underground and the fluids contained
within them comes with a high level of uncertainty and differences in scale,
resolution and amounts of data. For example, though seismic exploration data may
capture the vertical and horizontal extent of reservoirs, it has a rough resolution
on the order of dozens of meters, and is called soft data. Data from wells, on the
other hand, is obtained from rock samples from the oil–gas reservoir, and is on
the order of a few centimeters. In the case of a well into which a logging tool is
dropped, electrical, acoustic and density data around the well wall is obtained every
30 cm vertically. The interval of wells, however, is on the order of a few kilometers.
Consequently, the so-called hard data is very discrete. A geological strata model
is constructed by using these data appropriately and applying statistical methods.
Fig. 8 Flow chart of oil–gas field evaluation (ref. JOGMEC)
Estimation of the behavior associated with development and production is carried

out over several decades in time scale by numerical simulation of the oil reservoir
(see Fig. 8). From the mathematical point of view, various methods are also applied
to the proper use of these data and the evaluation of uncertainty included in the
result.
Related Fields During Exploitation/Development Stage and Utilization

of Data
– Exploitation/underground structure evaluation stage:

Three-dimensional seismic exploration data processing and analysis; uncertainty
evaluation using geostatistics for oil reservoir characterization; electromagnetic
data processing and analysis; oil generation/drift/sedimentation simulation; etc.
– Oil–gas reservoir reserves/production behavior estimation stage:
Flow behavior estimation using numerical simulation of oil reservoir (multi-
component, thermal recovery); application of genetic algorithm to oil reservoir
simulation/history matching; micro-flow simulation using lattice Boltzmann
method; methane hydrate melting/flow simulation; coupling of flow simulation
and rock compaction; sand behavior simulation; geomechanics evaluation simu-
lation; visualization of three dimensional data; etc.
60 H. Kawano
– Development/production stage:
Multiphase flow analysis using distinct element method; oil–gas–water manage-
ment process design; unsteady multiphase flow analysis in pipeline; offshore
production system mooring/riser tensioner behavior analysis; dynamic behavior
analysis of floating production system with dynamic positioning system on
board; etc.
Specific Cases
[Three-Dimensional Seismic Exploration]
One of the major achievements in petroleum and natural gas exploitation in recent
years is three-dimensional seismic exploration. In conventional two-dimensional
exploration, with data on a lattice grid at intervals of 500 m to a few kilometers,
data inside the gridlines is estimated using various interpolation methods, such as
spline interpolation or kriging. However, there is a high degree of uncertainty in
underground structure, which makes accurate evaluation difficult. Moreover, the
uncertainty is further affected by the surrounding underground structure, leading
to false images. Three-dimensional seismic exploration is implemented by placing
seismometers on a finer lattice with intervals of 12.5–15 m and recording the
reflected wave from every direction across the plane. This method can eliminate the
uncertainty accompanying interpolation, and increases the accuracy of underground
structure imaging by placing false images in the right position. However, capturing
the reflected wave on a plane increases the amount of seismic wave data recorded
considerably. A single survey conducted offshore from Abukuma over an area
of 70 km × 30 km by the three-dimensional geophysical survey vessel “Shigen”,
introduced by the Ministry of Economy, Trade and Industry (METI) in 2007 and
operated by JOGMEC, produced 12 terabytes of data. Optimization of parallel
computation and the algorithms used became essential to many processes, such as
noise reduction in the data, deconvolution to eliminate the effect of the hypocenter
(dynamite or air gun), and migration, which is a method for describing underground
structure based on seismic wave data (Fig. 9).
[Oil Reservoir Simulation]
A numerical simulation of the oil reservoir is used in the geological model

construction process to analyze changes in pressure inside the oil reservoir and
in the saturation rate of fluids (water, oil, and gas) during production. This
model is then used to optimize the production plan (see Fig. 10). In oil reservoir
simulation, physical phenomena governing the flow inside the oil reservoir, such
as viscosity, capillary tube pressure, gravity, and expansion of fluids and rocks,
are often expressed using nonlinear partial differential equations, which are then
solved computationally. The oil reservoir is modeled by dividing the space into
Fig. 9 Three dimensionalization of seismic exploration (ref. JOGMEC)
Fig. 10 Oil reservoir simulation model (ref. JOGMEC)
a finite number of cells, and the behavior of each portion of the reservoir is
estimated by solving the mass transport equations between cells for gas, oil, and
water phases using Darcy’s law. Multicomponent simulation is used for the more
complex interactions of carbon dioxide flooding and hydrocarbon miscible flooding,
treating the fluid as a mixture of several components and permitting component
62 H. Kawano
transfer between gas and liquid. In addition to calculating oil and gas phase flow
between cells, phase equilibrium calculations are performed. Variants of this type
of simulation include the thermal flooding model, which requires energy balance
equations, and the chemical flooding model, which expresses various chemical
phenomena. When it comes to multicomponent calculations with the order of a
few million cells requiring fast calculation and large memory, the calculation time
is large even with parallel computers. There are still problems with determining the
proper upscheduling, comparisons with actual production history, and automatic and
high-accuracy parameter calibration. A recent example of successful development
of a numerical simulator is a Japanese model that performed excellently in an
international competition for simulating the production of an unconventional natural
gas resource (methane hydrate). In this model, new routines were incorporated to
reproduce specific phenomena that had not been treated, such as melting gasification
of solid hydrate, rehydration, and consolidation of the unconsolidated sand layer and
changes in permeability.
Development and Application of Mathematics Expected in the Future in

the Petroleum and Natural Gas Industry
Applied mathematical problems concerning the aforementioned exploitation meth-

ods are as follows.
(a) Solving large matrix systems:
Problems in petroleum exploitation can be reduced to discretizing nonlinear
partial differential equation, building a large system of equations, and solving
them. In the case of physical exploitation, only a slight perturbation (error) in
the observed data may cause considerable changes in the output (underground
structure). To overcome this problem, we adopt the Tikhonov regularization
method and study stable solutions from matrix calculations. More efficient and
stable large matrix calculations are desirable.
(b) Inverse problems:
Optimization of inverse problem calculation is a major problem in migration
speed analysis, full wave inversion, history matching and well test analysis.
(c) Statistics is used to quantify uncertainty, and it would be desirable to improve
the accuracy using Bayes’ theorem and other mathematical methods.
These are the mathematical methods we, at JOGMEC, use in actual petroleum
and natural gas exploitation and development.
Company Profile of Japan Oil, Gas and Metals National Corporation

(JOGMEC)
In 2004, the Japan National Oil Corporation (JNOC) and the Metal Mining Agency
of Japan (MMAJ) were integrated to establish Japan Oil, Gas and Metals National
Corporation (JOGMEC) and ensure a stable supply of oil and natural gas, as well
as metal and mineral resources. Its primary roles are to implement exploitation
projects and technology development programs for promoting the exploration and
production of these resources, to provide necessary funding for exploration, and to
secure a stable supply of natural resources and energy through storage of petroleum,
natural gas and rare metals. JOGMEC also deals with protection of public health and
conservation of the living environment through mine pollution control measures,
and contributes to the sound development of metal mining.
A Brief Biography of Hirobumi Kawano, JOGMEC
Hirobumi Kawano was born in Tokyo in 1946. Following graduation from The
University of Tokyo with a B.D. in Economics, he joined the Ministry of Interna-
tional Trade and Industry, which was the predecessor of the Ministry of Economy,
Trade and Industry (METI). After holding director’s posts in the Petroleum Division
and the Basic Industries Bureau in the Agency for Natural Resources and Energy
(ANRE), he became Director-General of ANRE in 1999 until his retirement from
METI in 2002.
Kawano became an external director of Sony Corporation in 2003 and a senior
corporate officer of JFE Steel Corporation in 2004. He has been President of
JOGMEC since 2008.
His hobby is yacht racing. He is Vice President of the Japan Sailing Federation.
Iwane Laboratories, Ltd.
Waro Iwane
President and CEO
Iwane Laboratories
Mathematics in Our Company: What Does It Describe?
Whatever the nature of work or specialization, people often use mathematics in

real-world situations. This is also very true in our company, where activities are
performed based on realizing autonomous car driving and to manufacturing several
useful by-products which are developed during research on recognition systems.
When we talk about “recognition systems,” we always mean image recognition
systems. In analyzing three-dimensional space and trying to recognize what is
happening in the real world, we have to rely on mathematical methods. Some
examples that may be of interest will be discussed in the rest of this chapter.
From Visible to Computable—Image Analysis Technology

with Camera Vector
Camera Vector Value
A Camera Vector (CV) value has six degrees of freedom and is derived from the
three-dimensional (3D) position T and 3D posture R of an object. The important
thing is that a CV value can be obtained by statistically analyzing a two-dimensional
(2D) image that is taken by a camera. This process is called CV computation.
All 2D images that have CV values (which are called CV images) make it
possible for us to conduct 3D space analyses on every object appearing in the
66 W. Iwane
Fig. 1 Mechanism of omni-directional camera
images. Considering such images as perfect dioramas of the actual world, we are
allowed to identify 3D position or measure any size and distance between the
features.
When CV computation and CV images are applied to some technology for 3D
space analysis, we call such technology CV technology. CV technology is the
fundamental tool used in any solutions that our company provides.
Although CV computation can be performed with images taken by a normal
camera, we usually use an omni-directional camera to make the processing easier.
The picture taken by an omni-directional camera is of the 360◦ surrounding the
camera’s focal point (Fig. 1, left). We can then obtain a planar picture such as Fig. 1
(right) by applying an equirectangular projection to the spherical image.
Feature Point Tracking on an Omni-Directional Image

Using CV Technologies
Feature point tracking is a technique in which we analyze each 2D image frame

used in a motion picture, and then compare these images to find the unique feature
points.
The unique feature points can be found where the value of the cross-correlation
(convolution) function, shown in Formula ,2 is maximized. In computer process-
ing, the same method is applied discretely to image pixels, so we should find the
pixel (x, y) that maximizes Formula 3 (Fig. 2).
Calculating the CV Value
The actual position of a feature point in the 3D world is represented by a 3D vector

P. Further, the CV value of the nth frame is represented by a 3D position vector Tn
and a 3D posture matrix Rn (3 × 3 rotation matrix). P, Tn and Rn are all unknown.
Iwane Laboratories, Ltd. 67
Fig. 2 Feature point tracking
The only value we know is Pn , a 3D vector obtained by the polar coordinate

transformation of the tracking result of the nth frame.
We obtain Formula 4 under the condition that a projected point calculated from
P, Tn and Rn overlaps Pn . The sign ∼= means that both sides are equal in direction
rather than in length.
68 W. Iwane
Formula 4 holds at the feature points, of which there are usually hundreds.
Although the formula is very simple, thanks to the power of the notation it
becomes a huge system of equations when treated computationally. Moreover,
strictly speaking, there is no solution to the equations because there are more
equations than the unknowns and each of them has slight error. Instead, we process
the errors statistically and find a solution P, Tn , Rn , for which the error is minimized.
T R (P − T )
n n
∼
= Pn ···
4
Mathematics Bound Up with Business
The goal of our company is to achieve artificial intelligence (A.I.), and the
development of A.I. technology requires researching the function of the human brain
while describing the findings in mathematical language.
The fundamental ability required for A.I. is recognition ability. To recognize the
3D space around us, we must begin by describing this space mathematically.
Furthermore, to develop an A.I. recognition system, we must study recognition
in humans and describe it mathematically so that we can derive “solutions” from the
mechanisms.
Although we are only standing on the threshold of A.I. research, our next step is
to directly describe “intelligence” itself.
An important point is that the actual “intelligence” does not use mathematics,
although A.I. can be realized only with the aid of mathematics.
This reveals the uniqueness of mathematics, as well as its limitations.
So far, the “solution” has been obtained mathematically with the aid of a
computer. As a result, the 2D image can be analyzed as a 3D diorama, so that we
can understand the 3D space.
Aided by these methods, consequent images are analyzed automatically. Solving
the equations allows us access to 3D information extracted from the original 2D
images. We have developed the following products out of this research.
Virtual Geographic Information System in 3D (Fig. 3)
The Virtual Geographic Information System (VGIS) links together the CV images
and the Geographic Information System (GIS) digital vector maps. It is used for
improving road safety and in the operation and maintenance of roads. Since the
system allows 2D images to be used as 3D data, we can conduct 3D measurements
on the images, stabilize them, and add CG images to movies.
Fig. 3 3D-VGIS
WEBGIS (Fig. 4)
WEBGIS is a technology that allows streaming and dissemination of 360◦ videos

on web browsers in geographic relation to map data. One can perform real-time
measurements of 3D positions for a selected point in an image, so that it can
calculate the distance and flexibility to add tags. Software development kits (SDKs)
provided for third parties allow the user to use WEBGIS in conjunction with other
web mapping services to create a variety of web content.
RCG (Fig. 5)
RCG is a technology for constructing 3D geographical models from motion pictures.

It is realized by the high-precision frame tracking technology that is applied to
feature points. Realistic CG images are produced by dynamically applying suitable
textures to the 3D model.
70 W. Iwane
Fig. 4 WEBGIS
Fig. 5 RCG
Conclusion
In both our personal and professional lives, we cannot avoid scribbling letters and
writing sentences if we wish to express meaning. We can also use numbers, symbols,
formulas and equations to express universal truths, such as human thoughts and
natural phenomena, in an abstract manner.
Certainly, such thoughts and phenomena cannot be expressed without using the
power of mathematics. Since mathematical language is just as familiar as natural
language to us, however, we do not believe it has a special power. Rather, we tend to
see it as a trivial tool and sometimes do not even notice how vitally important it is.
However, we find it surprising that thoughts and phenomena themselves are made
from neither mathematics nor natural language.
Although thoughts and phenomena are not mathematics, they may be represented
by mathematics. It is important to remember that, for example, nature decides the
movement of an object without solving differential equations.
Our civilization could not have emerged without the invention of written
language, and we can say the same thing about mathematics.
It would be incorrect to think of mathematics as being used exclusively in
company activities. Rather, it is like air and water: both are indispensable for our
lives, but we rarely think about them.
When we encounter a difficult problem in human affairs, we often try to solve
it by asking questions in natural language, so that we are better able to understand
the essence of the problem. Similarly, when we wish to understand unknown natural
phenomena, we do the same thing with mathematical language.
When we take this approach, we use an everyday expression and say that we are
“solving equations.” However, although it may sound trivial, “solving equations” is
not a matter of indifference, but actually shows the unparalleled logical power of
mathematics.
Mathematics does not merely describe the problem in an abstract way, it allows
us to find a previously unknown “solution” from the abstract description. It is
surprising that the unknown can be transformed into the well known when we
succeed in describing the problem mathematically. This is a huge accomplishment
of mathematics that awakens our emotion.
As we have seen, running a mathematics-driven company is full of exciting
experiences. I am grateful for our human knowledge as well as our ancestors who
have developed mathematics to today’s level.
Acknowledgements I would like to thank Mr. Akifumi Arayashiki and Mr. Tomoyuki Wakakuwa
of our company, who provided valuable assistance while I was writing this chapter.
Company Profile of Iwane Laboratories, Ltd.
Iwane Laboratories was founded in 1979 and is headquartered in Sapporo. The

company is dedicated to researching and developing autonomous motor vehicles
72 W. Iwane
based on visual information processing technologies. It also manufactures many

by-products developed during research, including VGIS as described in this article.
VGIS is sold to the Ministry of Land, Infrastructure, Transport and Tourism (MLIT)
and its regional bureaus and road management offices, as well as expressway
companies located nationwide, having around 70% of the market share. Iwane
Laboratories also accepts proposals from partners worldwide for joint business
based on new technologies.
A Brief Biography of Waro Iwane, Iwane Laboratories
Waro Iwane was born in Sendai in 1943. After graduating from Yamagata University
with a B.D. in Physics in 1966, he started working as an assistant researcher at
the Research Institute of Applied Electricity (RIAE), Hokkaido University. His
achievements with RIAE include “Study on Information Processing Mechanisms
of the Human Visual System” and “Study on Band Compression Techniques in TV
Signal Transmission.”
He resigned from the institute in 1978 to start his own business. In the following
year, he established Iwane Laboratories and became the company’s President
and CEO.
Mathematics and I
Kaoru Yosano
Minister of Education (1994–1995)
Minister of International Trade and Industry (1998–1999)
Minister of Finance (2009)
Learning Math in English
After graduating from Azabu Elementary School, I went to Azabu Junior High
School. Today it is one of the most prestigious schools in Japan, but this was not
yet the case in my childhood. When I was in the seventh and eighth grades, I was a
poor student. One day my teacher called me out and said, “How long do you study
at home?” I said, “I work hard at school, so I don’t study at home.” At last, he said,
“You should study 15 minutes, at least.”
When I was an eighth grader, my father, who was a diplomat, was dispatched
to Egypt, and I was put into a British boarding school in suburban Cairo. From the
ninth grade, I started taking classes in English. Since the school rules were strict and
I could not understand English well, I had a hard time of it.
In those days, however, I made the wonderful discovery that math was easy when
taught in English. To solve a word problem in math, for example, you have to
ask yourself questions such as “what are the given conditions?” and “what is the
unknown value?” I could understand these processes more clearly than when I was
studying math in Japanese, because both math and English were languages that
emphasized the importance of clarity. After a short time, I became the best math
student in my grade.
I was not satisfied with only the official curriculum, so I sometimes bought math
textbooks myself and read them privately. It was a surprisingly big change for a
student like me, who had not studied at home in my home country. Among the
books I read in those days, for example, Do It Yourself Calculus gave me a good
understanding of calculus. I also became interested in physics and read George
74 K. Yosano
Gamow’s One Two Three . . . Infinity, in the eleventh or twelfth grade. While reading
these books, I experienced a lot of fun, as well as a vague sense of understanding of
the basics of math and science.
Hooked by Science
After spending 4 years in Egypt, I returned to Azabu High School as a twelfth grader.
I graduated from the Department of Law in The University of Tokyo, and then
started working for the Japan Atomic Power Company. Contrary to my expectation,
I was assigned to a research division in the engineering department. My supervisor
Ryukichi Imai was a graduate of the Department of Mathematics in The University
of Tokyo. He was the most intelligent man I had ever met. One day I complained,
“I do not understand what engineers are talking about,” to which he replied, “It is
nothing much to worry about. Anything will be clear if you study a little of it.” This
sparked my motivation, and I started learning nuclear power engineering.
I studied topics such as atomic fusion, fuel and, generally what was happening
inside the nuclear reactor. Some topics were beyond my understanding, as I lacked
the requisite mathematical background. For example, I did not understand the
diffusion equation used to describe the process of neutron diffusion. Despite these
difficulties, however, my interest in mathematics did not die. I studied hard on
nuclear power, and when I was tired of it, I enjoyed reading books about the history
of mathematics.
After I became a member of congress, I was interested in theories of elementary
particles and held onto the dream that I would settle down to study this field. Then,
shortly before my appointment as the Minister of Education, I started learning about
it from a book titled The Adventure of Quantum Mechanics (1991, Hippo Family
Club). This book contains many cartoons to help the reader’s understanding, and
was really a good book to start with, even though it did not describe every detail. So
I read other textbooks too, including Shin’ichiro Tomonaga’s Quantum Mechanics.
I was also interested in outer space and read works by Katsuhiko Sato, who was
known for studies of cosmic inflation. Although my understanding of quantum
mechanics and cosmology was very limited, I had a hunger for knowledge in these
fields.
“Elite Education” as a Means to Serve Society
There are two types of problem in the world. Problems of the first type can be solved
if they are attacked by a million people. The second type of problem can only be
solved by excellent people, regardless of how many there are. The proverb “Two
heads are better than one” is useless when it comes to the most difficult problems
such as proving Fermat’s Last Theorem. One million people of average intelligence,
Mathematics and I 75
working together, cannot reach the insight of one genius like Andrew Wiles. If we
want to apply the proverb, we have to alter it to “Two brilliant heads are better than
one equally brilliant one,” because brilliant discoveries only come from brilliant
brains.
I am not claiming, of course, that a few brilliant people can do anything. It is
true that a genius called Johannes Kepler discovered laws of planetary motion in the
17th century, but this discovery depended on the many anonymous people who kept
track of planetary movements each night and added to the huge amounts of data.
The best example is Tycho Brahe, a scrupulous scientist whose data led to Nicolaus
Copernicus’ heliocentric ideas, as well as to Kepler’s Law. Without such people, we
would have had neither Kepler nor Wiles.
To put it another way, there need to be a certain number of exceptionally brilliant
people for the efforts of ordinary people not to be wasted. If there was a revival of
so-called “elite education,” this would not be for the “elite” people themselves but
for society in general. In other words, we should raise elites and make them serve
the ordinary people.
I strongly agree that education should be given to all people equally. However, it
is also true that every child has both strengths and weaknesses. Some can run fast
while others can study well. We have to develop the abilities of talented children for
Japan’s science and economy to progress further.
Keep Learning
One of the reasons why I study math and physics is that I want to keep my brain
active as I get older. Mostly though it is because I have a hunger to understand how
the world is constructed. I am learning out of curiosity.
Math is quite fun when told as a story. For example, one day the following
question popped into my mind: “How did humans discover the general solution of
quadratic equations?” After reading a history book, I learned that it was discovered
in Arabia. The book also explained that the general solutions of cubic and quartic
equations were found in Italy, but that quintic equations were proved not to be
solvable by radicals. It is purely for fun to know such stories, so while some people
say that math is boring and has nothing to do with our amusement, I do not agree.
Sercrets of the World
A tanka poem is engraved on my grandfather Tekkan Yosano’s gravepost: “Having

seen nearly all the secrets of the world/now I see the vault of heaven.” I do not know
the real concept behind the poem, and it might be exaggerated for a poetic effect.
What I can do is simply wonder if humans can see “nearly all the secrets of the
76 K. Yosano
world.” Why do I think so differently from my grandfather? I suppose it is because

I have been familiar with math and science for a long time.
For example, consider electromagnetic waves. Humans are able to see so-called
visible light, which is a certain range of wavelengths within the electromagnetic
spectrum. If we want to see the infrared-, ultraviolet- and X-rays, we have to be
equipped with scientific knowledge and devices. We can also observe neither the
behavior of electrons nor of the interior of the human body without making use of
electronic or scanning microscopes.
Other examples are Einstein’s theory of relativity, which reveals that mass is
equivalent to energy, and quantum mechanics, which argues that contradicting
statuses might coexist. These discoveries were contrary to human intuition, and have
changed our understanding of nature and widened our mindsets.
In spite of such huge advances in science, however, we know only a small portion
of physical phenomena. If we were unable to use scientific knowledge and could
depend only on our five senses, our understanding of the world would be even more
limited.
Knowledge of science developed in me the belief that most things in the world
are beyond human understanding. In addition, as a politician, I came to think that
quantum mechanics, in which contradictory statuses might exist together, was a
good analogy for the complications of politics and other human affairs: even though
they are chaotic and seemingly contradictory, it is unnecessary to think that they are
unacceptable.
Having learned math and science, I do not believe that I can know “all the
secrets.” Instead, what I can do is to know as many secrets as I can, and that is
what I try to do.
A Brief Biography of Kaoru Yosano
Kaoru Yosano was born in Tokyo in 1938. Following graduation from The Univer-
sity of Tokyo with a B.D. in Law, he joined the Japan Atomic Power Company. After
working as a secretary to Yasuhiro Nakasone, who would later become the Prime
Minister, he was in 1976 elected for the first time to the House of Representatives.
Known as one of the most distinguished policy experts, he held prominent
positions such as the Minister of Education in the Murayama cabinet, the Minister
of International Trade and Industry in the Obuchi cabinet and the chairman of the
LDP’s Policy Research Council in the Koizumi cabinet.
He was the Chief Cabinet Secretary in the Abe cabinet, and in the Aso cabinet
tripled as the Minister of Economic and Fiscal Policy, the Minister of Finance and
the Minister of Financial Services.
He is a grandchild of the noted poets Tekkan and Akiko Yosano.
Mathematics for Industry: Principle, Reality
and Practice, from the Point of View of a
Mathematician
Masahiro Yamamoto
Professor,
Graduate School of Mathematical Sciences,
The University of Tokyo
We describe the principles of mathematics for industry and discuss schemes for
promoting activities. Then we present two cases of the practice of mathematics for
industry and argue on a bilateral collaboration style for effective joint projects by
mathematicians and industry.
Introduction
Originally mathematics was not only aimed at abstraction but also was applied to
solving real problems, and it has been developed in both respects. We can understand
that if we recall several prominent mathematicians such as Archimedes (287–217
BC) and J.C.F. Gauss (1777–1855). They worked for both abstract mathematics
and real applications of mathematics. In particular, Gauss invented the least-squares
method, which is a major numerical method nowadays for various problems.
The least-squares method is the basis of the Tikhonov regularization, which is a
fundamental method for inverse problems as explained in section Case Study I: Risk
Management and Innovation for Operation of a Blast Furnace.
Reviewing mathematical history, we can recognize that great knowledge of
mathematics, which was originally exploited only for itself and not necessarily
aimed at applications, has eventually played essential roles in other disciplines, as
in quantum physics and the theory of relativity. Thus it is natural to expect that
mathematics can work for industry. This is “mathematics for industry”.
78 M. Yamamoto
In this chapter, the author describes his views concerning:

1. How mathematics can work for industry.
2. What the status of mathematics is for industry in Japan.
3. What we should do in order to develop mathematics for industry.
Then we explain inverse problems as mathematics for industry and two examples
of success.
Advantages of Mathematics for Industry, Difficulties

for Activities in Japan and Short-Term Policies
For solving problems in the real world, not only a knowledge of mathematics is
useful, but also the mathematician is meaningful: the mathematician uses specific
ways of thinking such as abstraction, generalization, and extraction of truth from
various realities to crystallize to a simple statement as a theorem. Moreover, in
various applications, mathematicians do not merely apply knowledge in an existing
specific area of mathematics, but can build up a new theoretical system of math-
ematics for solving concrete problems. Therefore, these intrinsic characteristics of
mathematicians may frequently yield breakthroughs for interdisciplinary projects.
In particular, we can expect innovation in industry and more economic profits by
mathematics after proposals of problems from industries, because:
• The mathematical approach may be different from conventional cooperative
research, and we can find more fundamental solutions by mathematics which are
based on principles and theorems. The ways of thinking of mathematicians are
less dependent on existing hardware and methodologies, compared with those
of engineers, and can be expected to propose initiating ideas for flexible, non-
conventional, and innovative methods.
• With the help of the universality of mathematics, we can find general solutions
in various issues and themes which emerge across various industrial aspects.
Therefore we can solve several industrial problems simultaneously. Such help
can be offered only by mathematics.
• The activity range of mathematicians is totally international. With such inter-
national networks of researchers, if an industrial problem is suitably given to
mathematicians, then the mathematicians can invite members from their human
resources to place the right persons in the right places and to activate their
knowledge and accomplish the mission.
• Recently we can collect more data and we have to treat big data which increase
day by day. Among the collected data, noisy data or meaningless data may be
mixed in, and we should often select more meaningful data among the total data
set. For this, we can expect new mathematical models on the basis of geometry,
analysis, etc., which can distinguish relevant data sets from meaningless data and
make more effective analyses and designs. In particular, by more collaboration
Mathematics for Industry: Principle and Practice 79
among mathematicians and statisticians, effective analyses and designs can be

made by better interpretations of raw big data. In industry, it is more and more
important to exploit essential laws hidden behind tremendously large data, and
for modeling, mathematical methodologies provide hope. In section Case Study
I: Risk Management and Innovation for Operation of a Blast Furnace, we explain
a related successful example where an improvement of data interpretation is
yielding profits.
Moreover, as several cases show, collaboration with industry frequently brings
inspiration for new research fields in mathematics. In other words, mathematics
for industry is fruitful not only to industry but also to mathematicians. For
example, we refer to a joint research project on anomalous diffusion phenomena in
inhomogeneous media which was motivated by industrial and engineering interests
and is providing graduate students with seeds for writing mathematical papers. Here
we refer to Nakagawa et al. [3].
In spite of these expectations, in Japan there is still a small number of activities
of mathematics for industry, for example, manufacturing industries. On the other
hand, there are remarkable activities overseas: in Europe and China, for instance,
there are research organizations of mathematics for industry which regularly work
to deliver certain results, and such activities are well recognized and promoted by
academia, industry, and governments. Here, by academia we mean organizations
which conduct research not primarily aimed at profits. For a report on activities
related to mathematics for industry, see the OECD report in 2009:
http://www.oecd.org/dataoecd/31/19/42617645.pdf
Thus, by the worldwide level of excellence of Japanese mathematics, we should
naturally expect more mathematics for industry in Japan.
Next we can analyze possible reasons for the current status in Japan as follows.
1. There is a deficiency of know-how on industry–university cooperation: a diffi-
culty in building constant credibility between industry and universities.
2. There is a small number of mathematicians as coordinators. Here, by a
“coordinator” we mean a person who can understand not only the requirements
about time-frame and desired outputs from industry, but who also can exercise
strong leadership within research groups and networks, which can build bridges
between industries and mathematicians. The same is happening on the industrial
side. As a result, research projects tend to be carried out with a unilateral
direction and do not last for a long period of time.
3. The timescale mindset on the university side is much slower than on the industry
side after proposals of certain missions from industry. The university side is too
much inclined to regard a topic put forth by the industry side as a part of the
educational and recruitment activities in graduate schools. Mutual understanding
about timescales and purposes by both sides is missing.
4. There is no flexible system on the university side which enables mathematicians
to solve missions presented by the industry side.
80 M. Yamamoto
5. In Japan, there are not many post-doctoral mathematicians who are competent
in theory, programming, and computation. That is, in all three, a great number
of students at departments of mathematics are not well motivated to apply
mathematics to industrial missions, although their mathematical abilities are very
high.
6. The actual results delivered through industry–university cooperation have not yet
been sufficiently evaluated on the university side. That is, there is a gap between
evaluation standards in the university and those in industry: the standard of
evaluation within the university is based on publications in international journals,
while the standard within industry is usually by patents or real use. Moreover, it
may not be possible to publish some achievements in view of confidentiality
requested by the industrial side, so that the achievements by mathematics may be
invisible.
These difficulties may exist everywhere in industrial countries, but in the
countries where these are better solved, activities by mathematics for industry are
conducted more systematically to yield constant outputs. One definite solution is
organization of institutes of mathematics for industry.
The author proposes short-term policies for overcoming these difficulties from
a general viewpoint, which will be conducted appropriately by the presence of
institute of mathematics for industry:
• Assigning mathematician–coordinators in the above sense.
• International–interdisciplinary networks conducted by the coordinators in order
to solve missions within a requested time limitation, by inviting appropriate re-
searchers on appropriate spots. This network is needed for overcoming deadends
or bottlenecks which may appear during limited-duration research.
• Pool of young talent: We should always provide suitable opportunities for
graduate students in order for them to understand problems in the real world,
as in industry.
• Understanding of time frames and scheduling by both the university side and the
industry side.
Example: It takes a huge amount of time to translate real-world problems into
mathematical language. Even for bright mathematicians, it may take a substantial
amount of time for solving this issue.
• Understanding by the industrial side of the value of fundamental research in
various mathematics, because we cannot predict which mathematics will be
useful.
Our Current Efforts
In the previous section, we briefly stated policies for overcoming the difficulties.
Needless to say, for their realization, we need more concrete efforts. As one
example, we mention current efforts at Graduate School of Mathematical Sciences
of The University of Tokyo by the Global COE Program ”The Research and
Training Center for New Development in Mathematics,” which started in 2008. The
GCOE Program contains the following activities resulting from the financial support
and human resources of the program.
1. Course of Mathematics in Industry
Since 2009, we have offered a course of mathematics in industry every year for
graduate students, and a student can gain credit units from attending. The course
is shared with about seven speakers from IT, insurance, finance, and manufacturing
companies. The purpose is for graduate students to see how mathematics is applied
in various actual worlds of industry and to make better plans for future career paths.
2. Guided Tours to Industrial Sites, and Discussions
Since 2011, we have organized guided tours to plants and institutes of Kao
Corporation and Nippon Steel Corporation. The purpose is not only to see the plants,
but also have the students present overviews of their own research to members of
the visited companies, which may be useful in considering future possible industrial
collaboration.
3. Study Group for Solving Problems Posed by Industry
As one difficulty for industrial collaboration of mathematicians, we pointed out

many gaps (e.g., timescales) in research styles between the university and industry.
It may take time to fill up such gaps, and we believe that the study group is
one effective way. That is, the study group aims at (a) providing mathematical
solutions of problems in industry and (b) finding seeds and fostering for possible
joint research projects.
The study group originated in 1968 at University of Oxford, where the main
organizer was Professor John Ockendon. The program of a 1-week study group
is as follows: (a) On the first day, the industrial side presents their problems to
participants, who consist mainly of graduate students. (b) The participants choose
the problems freely according to their interests and, divide into several teams, and
try to reach solutions with presenters from industry. (c) Graduate students make
reports on the achievements obtained during the workshop period.
82 M. Yamamoto
The expected outputs are

• Intensive work by mathematicians towards realistic solutions.
• Participating graduate students are offered meaningful occasions where their
mathematics may solve concrete problems from the real world, and as a
consequence can be provided incentives for their own mathematical research
and obtain deeper understanding of mathematics. Thus the study group can be
considered as one program in a graduate course.
• Experiences for graduate students: Even in the case where a final report
cannot give sufficient solutions, the students can have valuable experiences in
interdisciplinary collaboration, and the company side can make the problems
clearer from a mathematical viewpoint.
• Smooth entrance to mathematics for industry: With a strict joint research and
development agreement, graduate students and other participants from academia
must produce outputs on a short timescale and usually hesitate to join, while
industry can not openly search through a wide range of mathematicians.
• Suitable platform for smooth collaboration: It can be expected that at an open
workshop, the company may be able to provide missions to a wide range
of mathematicians and gain various knowledge and methods to exploit novel
solution methods. This is impossible within closed projects subject to strict joint
research and development agreements.
There are several examples where more exclusive joint research projects have
started after study groups.
Study groups have continued to be organized in England, China, Canada,
Australia, and New Zealand as far as I know. Taking into consideration the
above-mentioned status in Japan, we think that the joint organization by plural
universities is more effective. Since 2010, The University of Tokyo and Institute
of Mathematics for Industry of Kyushu University have jointly organized study
groups three times (in addition, twice was solely by The University of Tokyo): the
first half of each study group was at Kyushu University and the latter half was at
The University of Tokyo. Moreover, from 2012, it has been supported partly by
the Unit for Mathematical Innovation of Basic Research Promotion Division of
Research Promotion Bureau of Ministry of Education, Culture, Sports, Science and
Technology in Japan (MEXT).
Please refer to the following websites:
http://sgw2011.imi.kyushu-u.ac.jp/
http://www.ms.u-tokyo.ac.jp/gcoe/activity/SG.html
http://sgw2012.imi.kyushu-u.ac.jp/
4. Research Internship
Since 2011, two graduate students have stayed 2 or 4 weeks at Advanced Tech-
nology Research Laboratories of Nippon Steel Corporation (at that time) and
worked for mathematical solutions and formulation of real problems posed by the
company. The internship for recruitment is common but this internship is aimed at
mathematical activities by graduate students in company. We expect development
of this program in the next year.
As Professor Masato Wakayama writes in this volume, there are comprehensive
activities at Kyushu University organized by Institute of Mathematics for Industry
and the GCOE Program ”Education and Research Hub for Mathematics-for-
Industry.” Also at Graduate School of Mathematical Sciences of The University of
Tokyo, we are substantially expanding related activities. The combination of these
multiple activities should be useful for promoting mathematics for industry involv-
ing graduate students and conducting an open platform for industrial collaboration
with mathematicians.
Applications of Inverse Problems for Industry
The author has been working on inverse problems. By inverse problem, we mean
a problem of determining a cause or detecting invisible objects by available data.
For example, medical diagnosis and non-destructive testing are inverse problems.
Also one can schematically understand it in Fig. 1. We can interpret “input” as
a cause or inaccessible object, and “output” as results or measurement data. The
“model” describes the law which governs the phenomenon and connects input
to output, and is often given by a differential equation. The inverse problem is
the determination of an input and/or a system by an output, while the forward
problem is the determination of output for a given system and input, which is
typically a prediction of the future state of the system. Moreover, a problem of
designing the input or system in order to realize the output, is usually called a
control problem. The inverse problem and the control problem are based on the
same concept. A tremendous number of problems in industry can be considered as
inverse problems. Since the inverse problem is a determination problem of some
states by incomplete data, the inverse problem commonly has intrinsic instability.
That is, even if noises in measurement data are small, huge deviations may occur in
resulting solutions reconstructed by noisy data.
Already we have considered the general advantages of mathematics for in-
dustry and here we add other advantages of mathematics of inverse problems
for industry.
• By industrial inverse problems, one can expect nice cost performance by the
participation of mathematicians.
In the inverse problem discussed in section Case Study I: Risk Management
and Innovation for Operation of a Blast Furnace, the demanded mission is to
input model output

Fig. 1 Input–output relation
via model
84 M. Yamamoto
create methods for data interpretation yielding reasonable estimation of interior

heat-flux, where data have been already collected. The issue is concerned with
extraction of essential information from big measurement data. For that inverse
problem, in principle, we need not prepare new equipment for measurement and
can rely on already collected data.
• Mathematical specification may be useful for the succession of technology. There
are many masters’ skills which were obtained on the basis of tradition and
experience, and without such skills one cannot manufacture products of high
quality. In order to pass such traditional skills over generations, it is useful to try
to clarify them as quantitatively as possible. Therefore it is significant to learn
such skills not only by experience, but also by clarifying the governing laws
behind phenomena. For this, mathematics is a reliable tool and one example is
given in section Case Study II: High-Quality Laser Hardening by the Inverse
Problem.
• In the view of mathematics, one can comprehensively review and reconsider
specific solution methods which have been used individually for concrete
problems in the real world. Mathematics offers general theories, and so under
unified principles, we can uniformly treat several problems which look very
different and were inclined to be considered separately in different sections of
industry. If one can introduce suitable ways of mathematical thinking, then we
can expect to treat several problems simultaneously and share gained knowledge
and techniques over one working section, which may yield savings of cost
and time. For example, because of the instability of the inverse problem,
we need special techniques such as the Tikhonov regularization stated below
and common methodologies for overcoming the proper instability to obtain
reasonably accurate solutions within the noise level. Such instability and relevant
methods for the inverse problem are well studied mathematically. Therefore,
given a concrete inverse problem in industry, it is more reasonable to first clarify
the problem in view of the general mathematical theory and then take into
consideration specific factors of the given problem, rather than to consider the
given problem ad hoc according to different situations in industry. In reality, it
seems that frequently researchers in different fields have been working separately
on essentially the same problems which are seemingly different, but in the view
of mathematics, one can uniformly organize such research to yield more effective
output.
Case Study I: Risk Management and Innovation

for Operation of a Blast Furnace
A blast furnace controls a basic process for the iron-making process, which produces
molten iron from sintered ore and coke through the deoxidizing process. The
inside temperature distribution is highly nonhomogeneous and is around 1,500 ◦ C.
Fig. 2 Cross section of blast furnace: raw data by thermocouples, reconstructed heat-flux [2]
A large-size blast furnace is about 100 m high in total and produces approximately
12,000 tons per day and, moreover, should work without any break down for more
than 15 years.
We are concerned with the basement of the furnace, which is covered by special
bricks, and the brick directly touches the molten iron (see Fig. 2a from Nakagawa
and Yamamoto [2]). The brickworks are expected to withstand the high temperature
and are designed to protect the plant and avoid damage to the furnace. It is extremely
necessary for us to estimate the heat-flux on the inner surface of the brickworks
because the interior heat-flux can be a good index for a secure controlled process.
If the heat-flux behaves irregularly, then one has to reduce the activity level of the
furnace. By considering the daily production output (i.e., about 12,000 tons per day),
the reduction of the activity level implies significant economic loss, and should be
avoided as long as safety is assured.
However, direct observation of the interior heat-flux is impossible owing to the
size and structure of the furnace and the high temperatures, so that it is possible
only to estimate them by temperature data near the bottom of the furnace which are
observed by thermocouples (Fig. 2a).
When irregular behavior of such temperature data is observed (e.g., Fig. 2b,
between 600 and 2,000 h) and, considering such an abnormal state as an omen
of worse irregularity, the furnace engineers conventionally carried out shutdown
operations for reducing the furnace activity (Fig. 2b from [2]: especially during the
86 M. Yamamoto
irregular period of measured temperature). Such unscheduled shutdowns suspend

the production of molten iron but must be done intermittently over a few months for
safety.
Therefore, important issues are:
• Can we reasonably estimate how the interior heat-flux behaves by available outer
temperature data?
• Is there an effective means of predicting such an abnormal period?
The first issue is concerned with what is called the Inverse Heat Conduction
Problem, and the second is related to risk management on the basis of an inverse
problem method.
Thus our mathematical task is the evaluation of heat-flux on the molten iron
surface by faraway data. We have the following technical issues:
1. The model equation should be a minimal mathematical model matching obser-
vation data.
2. We have to exploit a numerical method which is robust against errors in data and
can be handled on sites.
3. The numerical method should give a criterion for evaluating the interior status of
the blast furnace in order to maintain the efficiency of operation of the furnace
most safely and profitably.
Our work for this industrial mission is composed of the following four steps.
First Step: Relevant Mathematical Model
Inside the blast furnace, complex processes go with multiphases involving solid
and liquid phases under high temperature, and the modeling itself is a very difficult
problem. However, our main purpose is to find an index for a secure operation of the
furnace, not to know the inside process itself. Thus our strategy for establishing a
model equation is the choice of a minimal model explaining raw data of observations
by thermocouple. Indeed we simplify the shape of the concerned part of the blast
furnace (Fig. 2a) and choose a surprisingly simple one-dimensional heat equation:
ut (x,t) = α uxx (x,t), 0 < x < , 0 < t < T, (1)
where the constant α > 0 is the thermal diffusion coefficient, 0 < x < is the depth
variable of the brick, and x = and x = 0 correspond respectively to the bottom
end of the brick and the interface with the molten iron. We take finite differences
of measured data by two thermocouples located in the depth direction and we can
approximate the derivative ∂∂ ux (,t), 0 < t < T . Therefore our available measurement
data are
u(,t) = h(t), 0 < t < T, (2)

ux (,t) = g(t), 0 < t < T. (3)
Thus our inverse problem is formulated as

Inverse heat conduction problem. Determine
f (t) := −ux (0,t), 0<t <T
by g(t), h(t), 0 < t < T .

Here we note that an initial value u(x, 0), 0 < x < is also unknown.
Remark 1 (The Minimal Model). Here we choose a drastically simplified Eq. (1).
In fact, although the brick part is a three-dimensional object, we neglect other
directions of heat conduction and take only the direction in the depth. Actually
the area of the brick touching molten iron is two-dimensional, not one point
x = 0, and some part of such a brick face may be covered by solidified iron, and
another part may directly touch the molten iron. It is likely that such a solidified
iron layer melts down or increases in time by the time-evolutionary heat flow
from the deeper part of the furnace. For more complete modeling at the end of
the brick, such a heat flow from the deep part must be considered, so that we
must be concerned with the main reaction process in the huge interior part of
the furnace involving very many physical factors such as time evolution of solid–
liquid phases, convection and the highly nonlinear thermal diffusion coefficient, etc.
A possible governing equation may be represented by a free boundary problem for
the three-dimensional Navier–Stokes equations in multi-phase media coupled by a
heat equation whose coefficients depend on the unknown temperature distribution.
One can write it down mathematically, but even a fundamental theoretical study
such as the unique existence of the solution is difficult, and possible numerical
methods are tremendously time-consuming and also the justification of the possible
proposed method with real data is another difficulty. Thus making bold to neglect
all complexity, we have chosen a very simple model equation (1). Needless to say,
our choice of the model equation must be judged within the period of the research
project by how much our method can interpret the real data and yield a profitable
operation principle for the furnace. Thus our model has passed such examinations.
Second Step: Mathematical Subjects of the Inverse Problem
The inverse heat conduction problem belongs to a category of inverse problems for
partial differential equations.
The first mathematical subject is uniqueness. Uniqueness means: h(t) := u(,t)
= 0 and g(t) := ux (,t) = 0, 0 < t < T imply f (t) := −ux (0,t) = 0, 0 < t < T for
u satisfying (1). Uniqueness guarantees that there exists at most one heat-flux f at
x = 0 realizing given data g, h. We note that there may not exist such f . The proof of
88 M. Yamamoto
uniqueness is based on the classical theory of partial differential equations, and we

need intriguing estimates for solutions. The subject is one of the interesting and still
fruitful topics in the field of mathematical analysis. As for mathematical treatments
for inverse problems for the heat equation, we can refer for example to a survey:
M. Yamamoto, Carleman estimates for parabolic equations and applications.
Inverse Problems 25 (2009) 123013 (75 pp.).
The second mathematical topic is stability. Stability means that if g, h are small
in a suitable sense, then we can conclude that f is small. For a precise definition,
we have to specify the sense of the smallness, and for example we can choose the
1
2
L2 -norm · for measuring data and solutions: g = 0T g(t)2 dt .
This inverse problem is not stable, that is, even if changes in data g and h are
small, the deviation in reconstructed solution f (t) may be extremely large in the
L2 -sense. We can see the instability by the following example. Let n = 1, 2, 3, . . . .
and let us set
2
απ 2 n + 12
μn = .
2

We consider un (x,t) = ε ne−μnt sin n + 12 πx where n = 1, 2, 3, . . . and ε > 0 is
a parameter. We can directly verify that un satisfies (1) and hn (t) := un (,t) =
(−1)n nε e−μnt , gn (t) := ∂∂uxn (,t) = 0, 0 < t < T . Moreover fn (t) := − ∂∂uxn (0,t) =

−ε n n + 12 π e−μnt . We verify the instability with the L2 -norm. In fact, we can see
that
εn
gn + hn = √ 1 − e−2μnT ≤ √ ε , n = 1, 2, 3, . . ..,
2 μn π 2α
√ −2μn T
while fn = ε n n + 12 π 1−e √
2 μn
−→ ∞ as n → ∞. By choosing ε > 0 suffi-
ciently small, the solutions fn become very large as n → ∞ even if the data gn , hn are
small for all n, so that the inverse heat conduction problem is instable.
Instability can be understood intuitively by the fact that the heat conduction
process has a strong averaging effect of the temperature distribution: even if
the variations in f (t), 0 < t < T , are very large, the specific profile of f (t) is
greatly smoothened during conducting from x = 0 to the measurement point x = .
Therefore we have to reconstruct the original f (t), 0 < t < T by such smoothened
filtered data g(t) and h(t), and any large deviation in f (t) must be detected from
very small changes in observation data of g, h.
Remark 2 (The Restoration of Stability). It is known that if the uniqueness holds
for an inverse problem and the map from inputs to outputs is continuous, then we
can restore stability provided that solutions are restricted to a suitable bounded set.
In the above case with the L2 -norm, we can more precisely state it as follows. We
define a space L2 (0, T ) by
T

L (0, T ) := g;
2
g(t) dt < ∞
2
0
1
T 2
with the L2 -norm g := g(t)2 dt . Let M ⊂ L2 (0, T ) be a compact subset.
0

For example, as M we can choose g; g + dg dt ≤ C with an arbitrarily fixed
constant C > 0. Let un = un (x,t), n = 1, 2, 3, . . ., satisfy (1)–(3) with gn and hn . Then
we can conclude that

∂ un
(0, ·)
lim
n→∞ ∂ x =0
if limn→∞ ( hn + gn ) = 0 and
∂ un
(0, ·) ∈ M . (4)
∂x
This is a consequence of a general theorem in the topology which is also called the
Tikhonov theorem. This result implies that we can restore stability under an a priori
condition (4), and is called the conditional stability. For many inverse problems, we
can prove the conditional stability, and it is an important and attractive mathematical
subject to prove it with a concrete stability rate. We do not state the details.
Third Step: Robust Numerical Method Against Instability
We will give a sketch of our numerical method, assuming for simplicity that
g(t) ≡ 0. For the details, see Wang, Cheng, Nakagawa and Yamamoto [4]. The keys
are
• The Fourier method
• Discretization
• The Tikhonov regularization
For the initial/boundary value problem (1), (3) with given f (t) := −ux (0,t) and
initial value u(x, 0) = a(x), 0 < x < , we can use the Fourier method. By the Fourier
method, one can find the solution to the initial/boundary problem as superposition
of the trigonometric functions in x, which was established in 1822 by J.B.J. Fourier
(1768–1830) and is widely applicable to linear partial differential equations. The
Fourier method is classical and every undergraduate student must know it, and
thanks to our choice of a minimal model, we can apply it to write the solution as
follows. We set
n2 π 2
λn = α , n≥0
2
90 M. Yamamoto
and
2 ∞ −λn t nπ x nπ y 1
G(t, x, y) = ∑ e cos cos + .
n=1
Then we have
∞ t
u(x,t) = ∑ An (x)e−λnt − 0
G(t − s, x, 0) f (s)ds (5)
n=0
for 0 ≤ x ≤ and 0 < t < T in a suitable sense of the convergence, where

1
A0 (x) = u(y, 0)dy,
0

2 nπ y nπ x
An (x) = u(y, 0) cos dy cos , n = 1, 2, 3, . . . .
0
Now let us fix discrete times: 0 < t1 < · · · < tM = T . Our numerical task is to
construct approximate solutions of f (t) and a(x) := u(x, 0) by data g(t j ), h(t j ), j =
1, 2, . . . , M. We discretize the inverse problem. That is, first we truncate the series in
(5) to obtain
N tj
u(,t j ) ∼ ∑ An ()e−λnt j − 0
G(t j − s, , 0) f (s)ds
n=0
N j ti
= ∑ An()e −λn t j
− ∑ f (ti )
ti−1
G(t j − s, , 0)ds
n=0 i=1
ti−1

N j ti
= ∑ An()e −λn t j
−∑
0
G(t j − s, , 0)ds −
0
G(t j − s, , 0)ds f (ti ).
n=0 i=1
We set

i (x,t) = − 0t G(t − s, x, 0)ds, 0 < t ≤ ti ,
G
− 0ti G(t − s, x, 0)ds, t > ti
and
G i (,t j ), j ≥ i,
j,i := G j,i = 0, j < i.
G
Then we can rewrite the above approximate equation with respect to u(,t j ) to
obtain a system of linear equations:
PX = W,
where P = {P1|P2 },
⎛ ⎞
e−λ0t1 e−λ1t1 ··· e−λN t1
⎜ e−λ0t2 e−λ1t2 ··· e−λN t2 ⎟
⎜ ⎟
⎜ ⎟
P1 = ⎜ e−λ0t3 e−λ1t3 ··· e−λN t3 ⎟
⎜ ⎟
⎝ ··· ··· ··· ··· ⎠
e−λ0tM e−λ1tM ··· e−λN tM
⎛ ⎞
1,1
G 0 0 ··· 0
⎜G ⎟
⎜ 2,1 G2,2 − G2,1 0 ··· 0 ⎟
⎜ 3,2 − G3,1 G3,3 − G3,2 ⎟
P2 = ⎜ G G ··· 0 ⎟
⎜ 3,1 ⎟
⎝ ··· ··· ··· ··· ··· ⎠
M,1
G M,2 − G
G M,3 − G
M,1 G M,2 M,M − G
··· G M,M−1
and
⎛ ⎞
A0 ()
⎛ ⎞ ⎜ A1 () ⎟
u(,t1 ) ⎜ ⎟
⎜ ··· ⎟
⎜ u(,t ) ⎟ ⎜ ⎟
⎜ 2 ⎟ ⎜ ⎟
⎜ ⎟ ⎜ AN () ⎟
W = ⎜ u(,t3 ) ⎟, X =⎜ ⎟.
⎜ ⎟ ⎜ f (t1 ) ⎟
⎝ ··· ⎠ ⎜ ⎟
⎜ f (t2 ) ⎟
u(,tM ) ⎜ ⎟
⎝ ··· ⎠
f (tM )
By solving this linear system, we can obtain f (ti ), 1 ≤ i ≤ M and Ak (), 0 ≤ k ≤ N.

However, the coefficient matrix P is ill-conditioned, that is, P causes instability in
solving the linear system even if W is polluted with small errors. This is a natural
consequence of the original instability of the inverse problem itself.
In general, for the inverse problem, if we choose very fine discretization for the
numerical calculations, then the discretization may reflect the original instability of
the inverse problem very well and the resulting discretized problem is more difficult
for numerical computations.
For treating the instability of the inverse problems, we use the Tikhonov
regularization as a stable numerical method. This method was introduced mainly by
A.N. Tikhonov (1906–1993), who was a Soviet mathematician. It is considered that
he invented the regularization technique for solving various real inverse problems
such as image reconstruction.
The Tikhonov regularization is widely used nowadays and the main concept is to
replace the solution to PX = W by the minimization problem:
Min PX − W 2RM + γ X 2RM+N+1 (6)

92 M. Yamamoto
over X ∈ RM+N+1 . Here · RM denotes the Euclidean norm: (a1 , . . . , aM ) RM =

1
(a21 + · · · + aM ) 2 and γ > 0 is the regularizing parameter to be suitably chosen. If
γ > 0 is very large, then this minimization problem becomes similar to a trivial
problem Min X 2RM+N+1 over X ∈ RM+N+1 (i.e., the solution of this minimization
problem is X = 0). Therefore it is very stable but far way from the original problem.
On the other hand, if we choose γ > 0 to be very small, then the minimization
problem becomes similar to Min PX − W RM which is the same as the original
unstable problem. Thus the optimal choice of γ should be some medium value and
there is a choice strategy, which should be determined also in view of the accuracy
of observation data. We do not state the details, and roughly speaking γ should be
proportional to the noise level in measurement data for h, g. If not, we cannot obtain
the best numerical solutions. Thus with some tuning device in programming, we
have completed the numerical method.
Remark 3 (The Conditional Stability and the Tikhonov Regularization). In
Remark 2, we explain the conditional stability which asserts that we can restore
stability subject to some bounded set. This supports the Tikhonov regularization:
in the minimization problem, the second term on the right-hand side of (6)
automatically restricts X to a suitable bounded set where we can prove the
conditional stability and so the minimization problem is numerically well-
conditioned as a numerical problem. The analysis of the conditional stability is
not only an interesting mathematical subject, but also gives an important hint
for how to choose the crucial parameter γ for numerical methods for the inverse
problem. For some details, we can refer for example to
J. Cheng and M. Yamamoto, One new strategy for a priori choice of regularizing
parameters in Tikhonov’s regularization. Inverse Problems 16 (2000) L31–L38.
In order to establish a numerical method requested from industry, we have to take
into consideration the following issues:
• Fulfilling requests from real work: short CPU time, not special programs (e.g.,
without super-computers)
• Demanded accuracy versus realizable observation accuracy: Since the accuracy
of observation data cannot be improved drastically, the numerical method
requesting high-accuracy data is not helpful in practice, and we need a numerical
method producing accordingly reasonable results even with low-accuracy data.
• Theoretically consistent studies versus a convenient method which workers at
factories can use. Moreover, we should not sacrifice mathematical rigor.
Fourth Step: Implementation for Real Use
We have tested our numerical method for a miniature model built at a laboratory
of Nippon Steel Corporation and we have confirmed that our numerical method can
give better numerical results than other existing methods. Thus we can conclude that
our method is reliable.
Fig. 3 Collaboration with mathematicians [2]
This step essentially requested discussions conducted by the coordinators on both

the university and the industrial sides. This step especially cannot be done only by
mathematicians, and real joint work by mathematicians and industry is necessary
on the basis of mutual understanding. Bilateral collaboration is necessary and Fig. 3
[2] is a schematic explanation of that.
This is nothing but the understanding of one different culture by persons
coming from another culture. Usually such joint work is difficult, and without such
understanding, real collaboration by mathematics for industry is impossible. The
accumulation of activities as stated in section Our Current Efforts will be very
useful for overcoming possible culture gaps. Furthermore, this project has been
executed by an international team including mathematicians from Fudan University
in Shanghai.
We apply our method to real data by thermocouples over a time interval and
we reconstruct the heat-flux on the interface between the brick and the molten
iron (Fig. 2c from [2]). Our reconstruction results show huge oscillations of the
interior heat-flux before large changes of the observation data, while remarkable
oscillations are not observed during the period of observed high temperature of
the bottom part of the brick. This result can give an operation principle that we
should take suitable action such as reducing the power as soon as we can recognize
huge oscillation patterns in the reconstructed interior heat-flux. By our operation
principle, we can keep the efficiency of operation of the blast furnace maximal, so
that we can accomplish
• Reduction of production costs,
• Reduction of carbon dioxide emissions,
• Prolongation of blast furnace life duration.
94 M. Yamamoto
We note that one can never know the interior status, but the reconstructed interior
heat-flux can be used as an index for detecting indications of potential dangers and
that the reconstruction is by an inverse problem method.
Summary of Case Study I: Risk Management and Innovation

for Operation of a Blast Furnace
1. Our problem is an inverse problem and we cannot expect stability in the usual
sense. Therefore we have to apply a specific method called the Tikhonov
regularization.
2. For a realistic numerical method, we need other careful consideration for the
choice of discretization. In view of the intrinsic instability and accuracy of
available real data, we should not exploit any numerical methods which yield
highly accurate numerical solutions if data can be very accurate. In general, a
numerical method of high precision may well reflect the original instability and
is too sensitive to data, and the numerical results may not be reliable.
3. In many applications, available data are not of high precision. In the case of
the furnace inverse problem, the available data are not highly accurate because
the number of thermocouples is small compared with the size of the furnace.
Thus by such realistic situations, our numerical method should produce results
of reasonable precision even from medium-accurate data.
4. The Tikhonov regularization is a relevant method to produce reasonable nu-
merical results from not very accurate data. For it, the clever choice of the
regularization parameter γ is essential in view of the conditional stability, original
noise levels of data, and the accuracy of discretization. The key to the optimal
choice is the “golden mean,” that is, “Too much is as bad as too little.”
5. Joint research of the inverse heat conduction problem in industry is composed of
the three different steps:
(a) The mathematical part about the uniqueness and the conditional stability.
(b) The development of relevant numerical methods.
(c) The adjustment of the numerical method to real use.
Both mathematicians and industry are often inexperienced with part (c) (see
Fig. 3).
6. For the inverse problem, the most important mathematical subject is unique-
ness and then conditional stability. We note that usually the existence of the
input, realizing given output, is not necessary to be considered in the inverse
problem, because we are discussing a phenomenon which is clear and present.
Reconstruction by the numerical method on the assumption of its existence,
is much more important than the existence issue. For the well-posed forward
problem such as an initial/boundary value problem for the heat equation (1),
we note that uniqueness, existence, and stability are fundamental mathematical
subjects. A problem is called well-posed in the sense of Hadamard if we can

prove uniqueness, existence, and stability in finding a solution provided that
we suitably choose spaces of data and solutions, while an inverse problem is
called well-posed in the sense of Tikhonov if we can prove uniqueness and some
conditional stability assuming that solutions exist. Thus in the inverse problem,
the paradigm is different from the classical forward problem, and the inverse
problem should be considered within this paradigm.
7. The appearances of inverse problems are much more varied than those of the
classical forward problem and most of them are realistic and meaningful. The
universality of the inverse problem can be understood by the characterization of
the inverse problem in the first part of section Applications of Inverse Problems
for Industry, and the above methodology is widely useful.
Case Study II: High-Quality Laser Hardening

by Inverse Problem
As stated in section Applications of Inverse Problems for Industry, the inverse

problem is useful for the succession of traditional skills of artisans to the next
generations. Here we explain one example, which was a joint project of the author
with Professor Dietmar Hömberg of Weierstrass Institute for Applied Analysis and
Stochastics Berlin. As for the details, we refer to [1].
We consider an application for hardening of materials by thermal laser. This
treatment is indicated in Fig. 4 [1]. By p(t) we denote the laser power at time t.
Our task is to determine p(t) in the following sense. While the laser point moves
along a prescribed curve on a workpiece surface, the temperature uhs (t) in the hot
spot by the laser beam is measured by a pyrometer, which is a kind of thermometer.
We set up a controller called PID controller, which adjusts p(t) on the basis of
uhs (t) in order to realize a desired temperature h(t) at time t. In cases of uniform
thickness of the workpiece, it is sufficient to assume that the target temperature
Fig. 4 Pyrometer process

control of laser surface
hardening [1]
96 M. Yamamoto
h(t) is constant, which is below the melting temperature of the workpiece material
and is independent of t. However, when the thickness of the workpieces is not
uniform, a constant target temperature h(t) does not yield the uniform penetration
depth of the laser beam along curves under consideration. In other words, in a place
with little thickness, the constant power is too strong and the penetration is too
much, and one cannot have a product with uniform hardening. Therefore we need
to determine the laser power p(t) as a function in t in order to realize the set-up
temperature h(t) over some time interval 0 < t < T .
We formulate this problem mathematically, omitting the details. For a workpiece
with a flat upper face, by Ω we denote the flat face where the laser beam is radiated.
Thus we assume that Ω is a domain on the plane. In Fig. 4, the upper side of the
workpiece is the cross section of Ω in the depth direction and the laser is radiated
along a curve on the planar domain Ω . By x = (x1 , x2 ) we denote a point on the
plane with suitable orthogonal coordinates. The temperature evolution u(x,t), x ∈ Ω ,
0 < t < T is described by the following heat equation:
2
∂ 2u
ut (x,t) = ∑ ∂ x2 (x,t) + F(u(x,t)) + p(t)G(x,t), x ∈ Ω , 0 < t < T, (7)
k=1 k
u(x, 0) = 0, x∈Ω (8)

and
2
∂u
∑ ∂ xk (x,t)νk (x) = 0, x ∈ ∂ Ω , 0 < t < T. (9)
k=1
Here (ν1 (x), ν2 (x)) is the unit outward normal vector at x to ∂ Ω and (9) describes
no heat-flux through the boundary ∂ Ω . In (7)–(9), all physical constants are
normalized. We assume that the absorption of laser energy per unit area is modeled
by p(t)G(x,t) where G(x,t) is a normalized intensity profile and we suppose that
we know G(x,t) physically. In the case of a gas laser, the function G is of a
bell shape like a Gauss function. Moreover, high heating induces further effects
inside the workpiece such as solid - solid phase transitions and for the modeling,
we introduce a nonlinear function F in (7) satisfying suitable conditions. Now our
problem of achieving the uniform hardening is stated as follows:
Let γ (t), 0 < t < T , be a curve in Ω where the laser beam is radiated and let
h(t), 0 < t < T be a given target temperature. Then determine the laser power p(t),
0 < t < T , such that u(γ (t),t) = h(t), 0 < t < T .
We consider p(t) and u(γ (t),t) respectively, as input and output, as explained in
Fig. 1. This is a kind of inverse problem: we do not know p(t) in the heat source
term in (7) and we are requested to determine p(t) by means of desired output h(t),
0 < t < T.
We can solve this inverse problem by a modern theory of partial differential
equations called the semigroup theory, but we omit the mathematical details.
The semigroup theory has been developed since the 1940s mainly by Japanese
Fig. 5 Uncontrolled case:

temperature uhs in the
hot-spot of a laser beam (top)
and resulting hardening depth
corresponding to the fraction
of austenite a(T ) at end-time
T (bottom) [1]
mathematicians, one pioneer of whom was Professor Kôsaku Yosida, and its
systematic applications to inverse problems were done mainly in the Soviet Union.
We show numerical examples for the surface hardening of a workpiece with a
rectangular cavity (the lower part of Fig. 5 from [1] indicates a workpiece with
an originally created cavity shown in white). Figure 5 is a result of a numerical
simulation with constant laser power p(t) = p0 and constant velocity, that is, the
power is not chosen by an inverse problem method.
The top part of the figure indicates the temperature uhs of the hot spot on the
surface, while in the lower part the black shows a hardened surface layer (more
exactly speaking, this is the resulting fraction of austenite). This case is not good
because (a) as the graph of uhs (t) shows, the high temperatures above the cavity and
at the end of the workpiece lead to exceeding surface melting, and (b) the too-thick
hardening above the cavity may foster fatigue effects which may eventually lead to
destruction.
Next we show a hardening process with the inverse problem. In Fig. 6 [1], the
profile uhs is the temperature of the hot spot on the surface, and our control by
solving the inverse problem makes uhs lower above the cavity in order to avoid
exceeding surface melting. On the other hand, by usub we denote the temperature
profile close to the lower end of the hardening depth. Thanks to the inverse problem,
usub is uniform, which guarantees highly uniform hardening. Figure 7 [1] is a
photograph of a real product where the white layer near the upper surface is the
unform surface hardening part and the inside black part is of course the cavity. The
depth of the white layer does not depend on the locations, whether above the cavity
or not.
Am I permitted to write that Fig. 7 is evidence of one triumph of mathematics
for industry as well as the achievement described in section Case Study I: Risk
Management and Innovation for Operation of a Blast Furnace?
98 M. Yamamoto
Fig. 6 Controlled case:

temperature uhs in the
hot-spot of a laser beam and
usub close to lower end of
desired hardening depth (top).
The resulting hardening depth
corresponds to the fraction of
austenite a(T ) at end-time T
(bottom) [1]
Fig. 7 Result of a hardening

experiment utilizing the
computed temperature uhs as
the set-point for
machine-based control,
courtesy of Photon Laser
Engineering, Berlin,
Germany [1]
Conclusions
• We discussed the advantages of mathematics for applications to industrial

problems. Also we reviewed current difficulties in Japan and discussed our efforts
since 2009 for conquering such difficulties.
• We described two cases of industrial applications of mathematics which are for
real use: risk management of a blast furnace and high-quality laser hardening by
the inverse problem.
• The collaboration by mathematicians and industry should consist of complex
ingredients: mathematical theory, numerical computations, and real use. The
experiences can be true pleasures and further motivations to mathematicians.
Mathematicians need to remain aware of real work sites in collaborating.
• More graduate students should be concerned with mathematics for industry by
their level of mathematics. We should not change our ways of thinking about
mathematics and traditional research styles of mathematics, that is, pursuing
abstraction, and generalization, etc. We cannot a priori restrict the fields of
mathematics for solving problems from industry, and, as is stated at the end
of section Advantages of Mathematics for Industry, Difficulties for Activities in
Japan and Short-Term Policies, we cannot predict which part of mathematics can
be used. Mathematicians will be useful for industry provided that they maintain
a high level of mathematical knowledge and experience.
• Collaboration with industry frequently inspires mathematical research and even
creates new mathematical subjects.
• Bilateral collaboration between mathematicians and industry is highly desirable
and can produce a win–win relation. One scheme of such bilateral collaboration
is indicated in Fig. 3.
• Serious activities recently started in Japan for promoting mathematics for
industry. The above two greatly successful cases were based on more individual
grounds before those activities were begun, and so we can greatly expect
much more output from our current systematic efforts as described in the
previous section. On this occasion the author asks industry to remain interested
in mathematics and support the above-mentioned activities, and ask graduate
students to join the activities with a positive attitude to gain a wider vision.
Acknowledgements Professor Masato Wakayama (Kyushu University) and Professor Takashi

Tsuboi, Professor Yujiro Kawamata, and Professor Toshiyuki Kobayashi (The University of Tokyo)
gave valuable comments on the draft. However all the contents are the responsibility of the
author. Professor Yoshikazu Giga (The University of Tokyo) and Professor Toshiyuki Kobayashi
recommended that I write this chapter. The author thank all of them.
References
1. Hömberg D, Yamamoto M (2006) On an inverse problem related to laser material treatments.

Inverse Prob 22:1855–1867
2. Nakagawa J, Yamamoto M (2009) Bilateral practice of industrial mathematics in steel making
process. J Math Ind 1(B-9):157–163
3. Nakagawa J, Sakamoto K, Yamamoto M (2010) Overview to mathematical analysis for
fractional diffusion equations—new mathematical aspects motivated by industrial collaboration.
J Math Ind 2(A-10):99–108
4. Wang Y, Cheng J, Nakagawa J, Yamamoto M (2010) A numerical method for solving the inverse
heat conduction problem without initial value. Inverse Prob Sci Eng 18:655–671
A Brief Biography of Masahiro Yamamoto, Mathematician
Masahiro Yamamoto is a professor of the Graduate School of Mathematical

Sciences at the University of Tokyo. His research has focused on inverse problems
for differential equations. He has been involved also in various joint studies linking
industry and academia since 2000, covering topics such as risk management and
innovation for the operation of a blast furnace with Nippon Steel Corporation.
He is currently a member of the editorial boards of many journals including the
“Journal of Inverse and Ill-posed Problems” and “Inverse Problems.”
Importance and Unpredictable Effectiveness of
Mathematics in the Real World and for Industry
Masato Wakayama
Professor,
Director of the Institute of Mathematics for Industry
Kyushu University
It is known that there exist several fundamental mathematical conjectures which

remain unsolved (i.e., propositions that have not yet been proved but are anticipated
to be true). Interestingly and somewhat ironically, the process of searching for
proofs often raises interesting new problems that yield new insight into mathematics.
This in turn stimulates the development of new branches of mathematics.
Andrew Wiles’ 1994 affirmative proof of Fermat’s Last Theorem (conjectured by
Pierre de Fermat in 1637) and Grigori Perelman’s 2002–2003 proof of the Poincaré
conjecture (proposed in 1904) have resolved two key conjectures. Wiles kept his
attempt secret to avoid unhelpful speculation. Perelman declined to accept the award
or to attend the International Congress of Mathematicians to receive the Fields
Medal in 2006. Perhaps, the most important and famous among these conjectures
is the Riemann Hypothesis, which is closely related to the distribution of prime
numbers and was formulated by Bernhard Riemann in 1859. Finding a proof for it
has thwarted the efforts of numerous distinguished mathematicians for many years,
so that, even today, we still do not have confirmation of its validity.
On the other hand, since the mid-1990s, the dramatic improvements in computer
performance have forged a much closer linking of mathematics to science, tech-
nology and industry. As a direct result, there has been an unprecedented expansion
in the development of science, technology and industry with mathematics being
the language that guides and focuses the research. Consequently, today, society has
enhanced expectations about the development of new fields of mathematics and
about its increasing importance in all human endeavors including improving secu-
rity, understanding climate change, the control of biological invasion, preventive
medicine, etc.
102 M. Wakayama
Equally important is the impact of applications on mathematics. The origins of

much of mathematics can be traced back to attempt to understand and quantify
the world around us. For example, much of Newton’s research was motivated by
the quest to understand the stellar motion. The computer revolution represents
validation of Pythagoras’ insightful comment that “everything is number”. This
also reminds us of the words of Galileo Galilei “Philosophy is written in that
great book which ever lies before our eyes—I mean the universe—but we cannot
understand it if we do not first learn the language and grasp the symbols, in which
it is written. This book is written in the mathematical language, and the symbols are
triangles, circles and other geometrical figures, without whose help it is impossible
to comprehend a single word of it; without which one wanders in vain through a
dark labyrinth.”
In fact, it is known that the validity of the abovementioned Riemann hypothesis
would have a significant impact on future developments of information security,
especially cryptographic research, and that such developments would stimulate the
development of new fields of mathematics. It is therefore crucial not to under-
state the importance of the future role that mathematicians will play in enriching
of the world of mathematics and in increasing the opportunity for mathematical sci-
entists to conduct pioneering applied research in science, technology and industry.
Brief Historical Comments About Mathematics in Japan
Japan is developing cutting-edge technology based on high-level scientific research

and established technological prowess. It is a world leader in science and technol-
ogy, with the goals of both promoting domestic prosperity and contributing to the
international community. In addition, from an international perspective, the level of
mathematics research in Japan is high, with pure mathematics playing a central role.
However, it has been pointed out, by experts in Europe and the U.S. as well as in the
2006 report of the Japanese National Institute of Science and Technology Policy [1]
and the subsequent article [2], that the interaction between Japanese mathematics
and other scientific fields and industry is rather weak. This situation is also reflected
in the lack of researchers in mathematics and the mathematical sciences studying the
application of mathematics to problems of practical importance. Highlighting one
aspect of this problem, the above-cited investigation [1] states the following: “The
lack of researchers in industry with a background in mathematics is striking. Among
the research and development teams considered in the study, representing industries
in Europe, the U.S. and Japan, nearly 70% recognized the need for researchers with
mathematical backgrounds, and while in Europe and the U.S., this level of need is
almost satisfied, in Japan it is not, with approximately only 26% of the need being
filled, i.e. there is almost 40% gap from the Occident.”
Industrialized Western countries have been making an effort on a national scale to
promote and develop the application of mathematics and the mathematical sciences
to science and technology (as highlighted in the OECD report “Mathematics in
Importance and Unpredictable Effectiveness of Mathematics 103
Industry” compiled in 2007 and in the follow-up report “Report on Mechanisms

for Promoting Mathematics-in-Industry” [3]). This commitment has been made not
only with regard to the promotion of research in mathematics and the mathematical
sciences, but also with regard to education. In stark contrast to this situation, the
investment by the Japanese government in mathematics has been quite small. In fact,
until recently, even the importance of mathematics from a national perspective
was overlooked [1]. (Indeed, in the fourth governmental science and technology
master plan of the Prime Minister’s Council of Science and Technology Policy
issued in 2011, the “mathematical sciences” are specifically mentioned for the first
time. Moreover, at the beginning of 2011, the “Unit for Mathematics Innovation”
was established in MEXT, the Ministry of Education, Culture, Sports, Science and
Technology, Japan and subsequently a new committee “Committee of Mathematics
Innovation” was also established under the Council for Science and Technology.
These should help to improve greatly the situation in Japan.)
With the great increase in the capability of computers that we have seen
since the mid-1990s, it has become clear that a broad range of scientific and
technological fields, including the social sciences, depend greatly in an essential
way on mathematics. For example, the possibility for the acquisition and storage
of large amounts of data has led in many fields to rely heavily on access to modern
computer technology, and to thereby appreciate the importance of mathematics (and,
of course, statistics). The following are just a few of the countless examples of this
trend: number theory and algebraic geometry underlie the technology of information
security, including encryption, certification and noise reduction; statistics and
probability theory form the basis of research in finance and economics; optimization
and control theory are essential in the development of shipping, electrical power
transport and manufacturing processes; statistics and inverse problems employing
differential equations and integral geometry are important in the technology of
weather prediction, CT scanners, MRIs and the control of blast furnaces; discrete
geometry and statistical optimization are used in fields that rely on graphics, such as
engine design, entertainment and architecture; geometry and topology are of great
importance in materials science.
In the past, the focus in statistics was to determine how, from a small sample of
data, to draw inferences about the property of the entire population. Given the
phenomenal increase in the computing and storage power of computers and the
expectation that we are entering a new information age, it had been speculated for
approximately a decade that perhaps statistics would become unnecessary. Indeed,
the same had been suggested even for some aspects of mathematics. It is even being
suggested that, if our ability to carry out massive calculations becomes great enough,
mathematics would no longer be necessary. However, the fallacy in such speculation
has become abundantly clear. In the first place, statistical methods are absolutely
necessary in order to extract the meaning contained in large amounts of data. In fact,
rather than making mathematics obsolete, the present age of computers has made
mathematics and mathematical theory even more important, as it is necessary to
evaluate the data and results that computers generate and provide, to formulate
theories based on these results, and to insure the safe and proper operation of all
the related technology.
104 M. Wakayama
A few decades ago, before computing capabilities became so great, there were
many Japanese engineers and professors, in particular, in such fields as mechan-
ical engineering, aeronautical engineering, precision instruments engineering and
electrical engineering and, of course, physicists who could function as applied
mathematicians when visiting Western countries. In addition, in the past, the relation
between mathematics and engineering/technology in some sense holds via scientific
fields, in particular, physics. Today, however, with the development of high-speed
computing, the direct connection of mathematics with all fields of science and
engineering has increased strikingly, and mathematics has come to play the role
of both the language and the guiding framework for research in an unprecedentedly
broad range of scientific, engineering and technology fields. However, despite this
trend, it has been found that with the vast and rapid development of computer
software (mainly in foreign countries), the number of researchers in scientific
fields with a strong interest in mathematics has decreased greatly. Also, while this
trend is seen even in industrialized Western countries, it is more striking in Japan.
Given this situation, the training of young researchers with mathematical skills
necessary to solve real-world mathematical problems should be given a priority,
both domestically and internationally [1, 3].
In Japan, in the past, the education of graduate students in mathematics and
mathematical sciences took place primarily within the mathematics departments in
Faculties of Science. Also, before the shift in Japan, which was initiated by the
Government, to an increased role of graduate school education, the major focus of
such mathematics programs was the training of college instructors and professors
in mathematics. In those days, the number of students who entered Ph.D. programs
was extremely small, and even the number of people entering master’s programs was
very small, in comparison to other scientific fields. This trend continued even after
the shift to a greater emphasis on graduate school education. However, now, even
in mathematics and the mathematical sciences, following the government policy to
increase the number of people with Ph.D. degrees, there has been a large increase
in the number of students entering graduate school, including in Ph.D. programs.
This would seem to suggest that a balance has been attained between the potential
demand for mathematicians with industry, scientific or technological experience
and the number of graduates coming from Ph.D. programs in mathematics and
mathematical sciences, as specified by government policy. In relation to this, it
should be noted that, according to an article [4] in SIAM News [SIAM: the
world’s largest industrial and applied mathematics academic society, the Society
for Industrial and Applied Mathematics (USA)], in Brazil, which is experiencing
remarkable economic development, there are presently about 150 graduates from
Ph.D. programs in mathematics and applied mathematics per year, but this number
is insufficient to support the industrial development of that country [4]. For that
reason, plans are now being developed to increase the number of Ph.D. graduates
while maintaining the quality of their training. (For additional historical comments
about Japanese mathematics, research and education, see also [5].)
However, as mentioned above, the number of people with mathematics

backgrounds that are hired by Japanese companies is still quite limited. One factor
cited as a reason for this is the traditional view (which is particularly strong in
industry) that mathematics is not useful in direct application to industrial research
and development. (Actually, the importance of mathematics is not self-evident.)
This view of mathematics results unfortunately from a failure by broad levels of
management in industry to recognize the need for researchers with backgrounds
in mathematics. Another factor contributing to this problem is that, although the
number of students graduating with Ph.D. degrees in mathematics has increased
markedly, the curricula of such programs are still aiming at producing professors of
mathematics for university employment, and, as a result, researchers with a broad
range of skills and interests needed for employment in scientific fields and industry
are not being produced.
Given the present role played by mathematics and the mathematical sciences
internationally, and given the fact of how Japan is presently rather out of touch with
this reality, there is now a pressing need to reform graduate school mathematics
education in Japan, with the cooperation of other scientific fields and industry.
The aim of this new educational system should be to produce researchers and
technologists with a strong basis in mathematics including statistics who can carry
out work in a wide variety of settings within society and to increase the number of
university professors who, while pursuing deep research in their own fields, have
a strong interest in and commitment to developing stronger connections between
mathematics and society. It is clearly a two-way process.
The Casimir Spiral for Innovation
One of the intelligent ideas for promoting scientific research and technology
development was proposed by Hendrik Bruget Casimir (1910–2000). Casimir was
a Dutch physicist, a student of Paul Ehrenfest at Leiden, and an excellent leader of
industrial research at Royal Philips Electronics. He made fundamental contributions
to the foundations of quantum mechanics and solid state physics, and is also quite
well known in mathematics communities for the Casimir operators. He is actually
well known for predicting a true effect of quantum mechanics; i.e. even in the
seemingly emptiness of the vacuum between two metal plates an electromagnetic
force is generated that pulls the two plates toward each other. After the prediction,
now referred to as the Casimir force, it was necessary to wait for some 60 years
before the experimental verification was achieved by Lamoreaux [6] in 1996
with 1% accuracy. It is worth noting that, in order to make the prediction he
computed some divergence series, which is equivalent to computing the special
value of the Riemann zeta function ζ (s) at s = −3 (Leonhard Euler computed this
divergent series around 1749 before the explicit formulation of the notion of analytic
continuation) (Fig. 1):
106 M. Wakayama
Fig. 1 Euler’s moon and sun; functional equation of the zeta function
1
ζ (−3) = “1 + 8 + 27 + 64 + 125 + . . .” = .
120
From the fall of 2009, DARPA (the U.S. Defense Department’s Defense
Advanced Research Projects Agency) has launched a project encouraging scientists
to work on ways to manipulate this quirk of quantum electrodynamics and recent
research has already found that the manipulation of the Casimir force is a major
challenge.
It was already realized by the middle of the 1980s that innovation is not linear and
certainly not driven only by experimentation and theoretical research in science but
is also strongly driven through the exploitation of technology. This observation had
already been made by Casimir in the seventies at the time when the linear model was
still considered largely valid. Technology uses science with a time delay of about 10
years, he argued. In turn, science is driven by new development in technology. And,
both progress together. This is the so-called Casimir spiral and is a key component
of innovation today (Figs. 2 and 3). Of course, the spiral pitch of the scientific and
technological Casimir spiral differs considerably depending on the field (see [7]).
As described above, with the remarkable increase in the capabilities of computers
that has occurred in the last two decades, the era in which the contribution of
mathematics to technology is communicated only through physics and related fields
Fig. 2 1950s linear model for innovation
Fig. 3 Recent status for innovation
is over. Today, mathematics and mathematical research participate directly in the

development of new technologies. As a result of this trend, the possibilities for the
two-way interaction between mathematics and technology have increased greatly.
For that reason, it is necessary to consider the relation between the progress in
mathematics and science and that in technology—the so-called “Casimir spiral.”
A seminal example, the achievement of which required a great deal of patience, is
the development of the RSA [8] and other encryption technologies, which have
become very important for modern society. Such technologies rely heavily on
the discovery of exceedingly large prime numbers and their prime factorizations.
Even though there are infinitely many prime numbers and, as discovered by the
Pythagoreans, every integer has a unique factorization into a product of primes,
the discovery is challenging because the prime numbers do not follow a simple
predictable mathematical pattern which is directly exploitable.
A Waka for Vacuum Force
Inspired by the Casimir effect (which is also called the vacuum force), we composed
the following waka ( : a 5-line 31-syllable Japanese poem) by the standard
technique called the “adaptation of a famous poem .” It exploits the fact
−s −1
that ζ (s) is given by the infinite product ∏ p (1 − p ) over all prime numbers p.
108 M. Wakayama
Fig. 4 Interesting coincidence of numbers
One may notice immediately that this is actually an adaptation of the following
famous poem composed by Fujiwara no Toshiyuki Ason :
Fujiwara no Toshiyuki Ason (birth date unknown −901 or 907) was a middle
Heian waka poet and nobleman. This poem is the 169th of the 1111 poems in the
Kokin Waka-shu , which is an early Heian waka Imperial anthology,
envisioned by Emperor Uda (r. 887–897) and ordered by his son Emperor Daigo
(r. 897–930) in approximately 905. The number 169 = 132 is reminiscent of a prime
factor of 1859, the year of the Riemann Hypothesis’ birth (Fig. 4).
Mathematics for Industry
The meaning of the title “Mathematics for Industry (MfI)” is different from that
for “Mathematics in Industry”. That for the latter phrase is restrictive—it tends
to be identified with the actual mathematics that specifically arises in the daily
management and operation of manufacturing. The former, however, denotes a new
research field in mathematics that may serve as a foundation for creating future
technologies. This title/notion was born from the integration and reorganization of
pure and applied mathematics into a fluidic and versatile form capable of stimulating
awareness in industry of the importance of mathematics as well as responding to the
needs of industrial technologies. The history of this integration and reorganization
has led us to the belief that this basic idea will someday find increasing utility. (See
[9] for interesting and important observations about the application of mathematics
in industry.)
Institute of Mathematics for Industry
The Institute of Mathematics for Industry (IMI) was established in April 2011
as Kyushu University’s fifth research institute. IMI was specifically planned to
be an offshoot of the Global COE Program “Education and Research Hub for
Mathematics-for-Industry” (FY2008–2012). [MEXT promotes the Global COE
(Centers of Excellence) program which provides funding support for establishing
education and research centers that perform at the apex of global excellence to
elevate the international competitiveness of the Japanese universities in various
fields of study.]
The idea behind establishing IMI can be traced to the 21st Century COE Program
“Development of Dynamic Mathematics with High Functionality” (FY2003–2007).
The aim of this program—a precursor to our Global COE Program MfI—was to
provide a broad view about mathematics to doctoral students and to open new
career paths for them. This is being achieved through, for example, an introduction
into the program of long-term internships in industry and the establishment of
the Mathematics Research Center for Industrial Technology (MRIT, April 2007–
March 2011), which is one of the common-use education/research centers at Kyushu
University. Since its birth in 1939, the Department of Mathematics in the Faculty of
Science at Kyushu University has enjoyed a long tradition of respect for striking a
harmonious balance between pure and applied mathematics. This balance serves as
the framework for the organization and management of the IMI.
Kyushu University has instituted a Graduate School/Graduate Faculty System.
According to its rules, University (faculty) members, in principle, belong to a
Graduate School or a Research Institute. A Graduate School is an educational
organization (education body) within the different graduate school of studies. IMI
and the Faculty of Mathematics are the partner bodies responsible for the Graduate
School of Mathematics.
Beside the standard educational activities at the graduate school etc, the ac-
tivities of IMI include (a) promoting collaboration with industry, (b) organizing
international conference, workshops and study groups, (c) holding seminars for
industry-academia partnerships and mathematical tutorials, and (d) fostering under-
graduate, graduate and Ph.D. programs which deliver students with a comprehensive
mathematical understanding of the two-way linking between mathematics and
110 M. Wakayama
Fig. 5 Development of
mathematics
Real-World applications (the Casimir-spiral for mathematics (Fig. 5)). The insti-
tute has 27 regular faculty members and consists of three main divisions plus
one laboratory: Advanced Mathematics Technology; Applied Mathematics and
Fundamental Mathematics; Laboratory of Advanced Software in Mathematics. In
addition, in order to efficiently promote the above projects/activities, a Visiting
Scholars Division and a Partnership Promotion and Technical Consultation Room
have been established.
As the reader can understand from the organizational structure depicted above,
IMI is staffed by both applied mathematicians (specializing in such traditional
fields as numerical analysis, partial differential equations, integrable systems and
probability theory, as well as more modern fields, such as data encryption, nu-
merical simulations, statistics and optimization) and mathematicians specializing in
traditional pure mathematics fields (topology, differential geometry, function theory,
representation theory and algebraic analysis). Most of these members, including
those with backgrounds in pure mathematics, conduct collaboration in joint research
projects which involve researchers in industry. It is particularly noteworthy that
the pursuit of this type of research activity is in perfect alignment with the fact
that, as described above, cooperative research with industry is inclined to require

increasingly more investigation of fundamental (mathematical) issues. As we know
from history, in order for mathematics to meet the demands of society, it is not
sufficient to simply carry out the research that society and industry explicitly
request. Although mathematics research and the manner in which it derives results
is said to be of a deductive nature, considering the actual process of such research—
testing algorithms/calculations with many examples, trial and error investigations,
etc.—it becomes clear that such research has essentially an inductive nature. With
this in mind, when considering the present-day problem of developing methods
to recover information from extremely large data sets, it is clear that the need for
fundamentally inductive research based on or communicated through statistics is
increasing rapidly.
Given the situation as described above, from the perspective of the research
philosophy that constitutes one of the core principles on which research activity at
IMI is based, the goal for “Mathematics for Industry” can be formulated as follows:
“To develop a new research field of mathematics whose goal is to produce
universal technologies.”
More precisely, the goal of Mathematics for Industry (MfI) is to create, develop
and systematize a new field of mathematical research for the purpose of constructing
and developing universal technologies. In addition, a key goal of MfI is to foster
increased applications of such technologies, seeking a kind of feedback from these
technologies that can be used to advance mathematics.
The term ‘universal technologies’ requires some elucidation. Here, we use this
term to refer to technologies based on mathematics that are robust in the sense
that they possess the possibility to be of universal applicability in society for many
years to come. Of course, we do not envision such technologies to be useful forever.
Instead, we consider as our focus the development of technologies that will serve to
support society in fundamental ways for at least 30 years (and hopefully 100 years
or more). In addition, in the body of knowledge to which ‘universal technologies’
refers, we include technologies that apply to new mathematical concepts; these
technologies include the expression of such concepts in simple terms that make
clear their inherent nature of thinking and algorithms through which these concepts
can be implemented (Fig. 6).
[Examples of the cooperative development of universal technologies and mathe-
matics]
1. The formulation of Ito’s calculus by the late Dr. Kiyoshi Ito (recipient of the first
Carl Friedrich Gauss Prize in 2006) was the theoretical driving force behind the
development of the field of mathematical finance. This application, which even
Dr. Ito himself could not have foreseen, has, through the field of economics,
played a major role in shaping modern society on a global scale. Conversely, this
application has been led to research in the new field of stochastic integration and
stochastic differential equations, and this has resulted in great advances in the
field of probability theory itself.
2. Akaike’s Information Criterion (AIC) originated in control analysis with the goal
of formulating quality assurance for cement. Since that time, however, it has
112 M. Wakayama
Fig. 6 Lacing-up for science, technology and mathematics
fostered the development of the new theoretical field of information criteria. This
in turn has led to great advances in the field of statistics and its applications.
3. It was thought for several hundred years that the field of number theory had
no applications. However, this is no longer the case. Number theory is of
vital importance in modern information security technology, which is based on
cryptographic theory. The genesis of cryptographic theory lay in the observation
of the computational hardness of the factorization of integers into prime numbers.
The theoretical need provided by such information technologies including data
compression and code theory, has stimulated progress in number theory and
algebraic geometry, which serves as the foundation of the latter-mentioned
research fields. To keep pace with advances in these developments, both number
theory and algebraic geometry have indeed achieved a highly advanced level of
abstraction that certainly far exceeds the dreams of Gauss (widely considered
the greatest mathematician of all time). Thus, in a manner that no one could have
predicted, pure mathematics has realized a great application that took 2,500 years
for the making, and through feedback from this application, pure mathematics
has itself advanced. This is an archetypical example of the cooperative advance
of mathematics and technology.
4. The concept of a statistical (or an information) manifold was introduced for
the purpose of facilitating the insightful treatment of probability distributions
of data as a whole. The field of information geometry, which was developed

for the purpose of investigating statistical manifolds from the geometric point
of view, provided stimulation that led to big progress in both statistics and
differential geometry. Through this combination of the fields of Riemannian (and
non-Riemannian) geometry and statistics, new areas of mathematical research
have emerged. This kind of fusion could not have been foreseen prior to the
developments described above. These new areas of research are even applicable
to treating the foundations of optimization theory, and through this application,
there is now the beginning of a new field of mathematics involving Jordan algebra
(originated initially in the context of quantum mechanics), which had previously
been an object of study only in abstract algebra and later in harmonic analysis
and number theory.
5. Research and development of CT scanners and MRI machines used in medical
diagnostics have been mathematically formalized within a framework of inverse
problems. This has been realized through the expansion of the application of
such mathematical theory to fields of science and technology, including the study
of astronomy, earthquakes and blast furnaces. The kind of research applied to
these medical technologies is now making significant advances as an academic
field. These developments have led to the deepening of research related to areas
of algebra and geometry based on representation theory and integral geometry,
as well as research related to harmonic analysis and differential equations. In
addition, it has led to great advances in the theory of numerical analysis, and the
application of regularization and variable Hilbert scales techniques to the solution
of inverse problems.
6. In the study of problems in the fields of materials science (which involves the
development of new materials) and molecular biology (including pharmaceu-
tical research), “computational topology” has become a new area of research.
A primary concern of computational topology, as its name suggests, is the
formulation of efficient algorithms for solving topological problems, or using
topological methods to solve algorithmic problems occurring in other fields.
Further application of computational topology to materials science and other
fields has provided stimulation to topology itself, and even though it is no more
than a decade old, computational topology has thus contributed to the advance of
its mother field.
7. In their application to real-world problems, pattern recognition research, which
has been regularly in the news in our modern, information-based society
(for example, in regard to crime investigation, the entertainment industry, and
personal authentication) as well as the new field of pattern theory research
(whose goal is to formulate methods of understanding sound, human expressions,
and even human perception) exhibits a very strong dependence on probability
and statistics theory. However, in the analysis involved in such research, fusion
with functional analysis, Lie theory (infinite-dimensional representation theory),
algebraic geometry and differential equations are presently being investigated.
In fact, it is now thought that new fields of mathematics are necessary for the
114 M. Wakayama
advancement of this research. A recent book on “Pattern Theory” [10] by David

Mumford, the 1974 Fields medalist for research in algebraic geometry, provides
insight into this situation.
The above examples imply that, even putting aside the direct application of
mathematics to industry as it exists presently, the development of MfI as an
academic field, as are the cases of pure mathematics (principally, the advancement
of mathematics for its own sake) and applied mathematics (whose goal is to
solve the mathematical problems posed by society) represents a new approach to
mathematics. In fact, pure mathematics and applied mathematics should be regarded
as playing equally important roles—like the two wheels/axles of a car. This point
can be understood from both a historic perspective and the recent changes in
attitudes in both Western developed countries and developing countries. Also, it
is important to place emphasis on searching for the mathematical aspects of real-
world problems and the possible mathematical structure behind them, and hence to
begin research from a yet non-existent foundation, i.e. a foundation that has yet to be
included in the body of human knowledge. From this perspective, at IMI, we regard
industry as a treasure house, hiding many mathematical problems of relevance to the
real world. While strongly promoting the study of such problems, we aim to provide
education in doctoral programs with a foundation in mathematics for researchers
who will conduct research in various scientific fields and industry. In this way, we
aim to directly address the problem posed by the great lack of such researchers
in Japan (as compared to Europe and the US). We hope to develop IMI into a
research institute unique in the world, in which researchers in pure mathematics
and applied mathematics, including statistics and numerical analysis, have strong
daily interactions and carry out joint research. Below, we introduce a few examples
of such research/education activity.
(a) Study Group Workshop
Study Groups with Industry are an internationally recognized method of technology

and knowledge transfer between academic mathematicians and industry. It was
initiated in Oxford in 1968. From 2010, a short-term camp-style Study Group
involving Ph.D. course students, initiated on a nationwide scale, has the students
attempting to solve challenging problems which various industries are facing. This
is the first full-scale Study Group Workshop in Japan organized jointly with the
Global COE program “The Research and Training Center for New Development
in Mathematics” in the Graduate School of Mathematical Science at the University
Tokyo. Examples are easily found where statisticians along with pure and applied
mathematicians have collaborated to attack some given problems. For instance, a
problem that was proposed by the Railway Technical Research Institute in the Study
Group 2012 was solved in an almost perfect way as a result of a collaboration of a
statistician (Fig. 7), a representation theorist, and inverse problem and numerical
PDE specialists. (For more detailed discussion, see [11].)
Fig. 7 MfI Study Groups in Kyushu and Tokyo

116 M. Wakayama
Fig. 8 An actual example of Long-term Research Internship
(b) Long-Term Research Internship for Ph.D. Course Students
The Graduate School of Mathematics, Kyushu University, introduced in 2006 a

program “Long-term internship in Industry for Ph.D. course students” as a part of
the curriculum for the new doctoral program “Functional Mathematics” (Fig. 8).
This Long-term internship provides a joint-research opportunity for the students to
work for 3 months or more at industrial R&D laboratories. More than 50 students
have already completed the program. The Long-term internships have turned out
to be far more successful than expected. They have not only attracted attention
from domestic universities but also from people working in industry. Some students
produced patents, papers, positions in industry and even initiated new collaborative
research opportunities between enterprises and faculty members. After obtaining
a Ph.D. degree in Mathematics in Kyushu University, almost 15% have obtained
positions in academia, 60% in industry, and the remaining have post-doc positions
in several universities and public research institutes. The following figure illustrates
a practical progression of the internship.
(c) Joint Projects with Industry
The following table lists the joint research projects, coming from various industries,
which have been conducted at IMI (Fig. 9).
Fig. 9 Joints projects with industry
(d) Journal of Math-for-Industry (JMI)
We launched an electronic journal (but have published a hard copy once a year), the
Journal of Math-for-Industry in April 2009, which is published at the institutional
repositories. This journal JMI is dedicated to the broadening of the horizons of MfI,
and publicizes swiftly and internationally achievements in education and research
on MfI. JMI presents original research papers and survey papers with original
viewpoints in all scientific disciplines where mathematics and applications jointly
play a basic or seminal role. Articles by scientists in a variety of interdisciplinary
areas are published. Research areas include significant applications of mathematics
to, in and for industry, including feedback from industry to mathematics, which is a
new initiative for MfI (Fig. 10).
Concluding Remarks: Joint Research Between Academia and Industry,

and Fundamental Research in Japan
In Japan, it has been only about 10 years since the time that joint research
between academia and industry began on a significant scale (to this time, mainly in
engineering, agricultural and chemical fields). However, as seen in the three graphs
below, it appears that the number of such joint projects, the number of filings for
and acquisitions of patents, and other related measures of such joint research have
leveled off in recent years. According to the analysis carried out by the Ministry
of Economy, Trade and Industry, in the past, joint research between academia and
industry was not common in Japan, compared with Europe and the US. However,
because Japan had strong foundations for high-level research in both academia and
118 M. Wakayama
Fig. 10 Third issue of the

JMI in 2011
industry, it was not difficult to transfer the knowledge that already existed in the
academic world to industry and thereby achieve rapid technological development.
This situation has now existed for 10 years. However, today, we are grappling with
more fundamental problems. We have reached the stage where problems concerning
the real foundations of research must be investigated.
It is often pointed out, even outside of Japan, that after the Meiji Restoration,
and again after World War II, Japan made great efforts to catch up with the
world’s advanced countries, in particular with the US. Consequently, nowadays,
it is highly regarded as a country at the cutting edge of technology. Around the
1980s, Japan had gotten the unenviable reputation (as “Free ride on basic research”)
of embracing new proven national and international technology with resounding
success and turning it into an innovation with a Japanese emphasis. This could
not be, however, seen as so obvious. Actually, for instance, beginning around
the mid-1980s, Japanese companies did invest a lot in fundamental research and,
of course, basic scientific research was always greatly appreciated especially in
academia. Indeed, its economic success throughout history is directly related to
this inherent Japanese originality and creativity to be innovative. That success was
achieved through technological innovation. However, the research and development
style that has been used until this time (particularly, after mid-1990) is no longer
Fig. 11 The status of industry–academia–government collaboration in Japan. (The original infor-

mation by MEXT is available (in Japanese) at: http://www.mext.go.jp/a menu/shinkou/sangaku/
sangakub.htm.)
120 M. Wakayama
sufficient. Indeed, for instance, before Japan became one of the world’s strong
leaders in technology, research and development in industry that had engineering
as its foundation, in particular manufacturing (monozukuri), was often sufficient,
but this is no longer the situation (Fig. 11).
As is happening in developed industrial countries around the world now, in
response to the demands and needs for the new technologies such as information
security, life science, etc., and even manufacturing, the new challenge for Japanese
industry is to embrace properly the linking of mathematics with industrial discovery
and innovation. Given this situation, we must also reconsider the importance of
mathematics education in Japan, while educating scientists and technologists with
strong backgrounds in mathematics and promoting a broad range of research by
these researchers. In this way, top-level mathematicians and mathematical scientists
as well as the fostering such researchers will become a vital part of the infrastructure
of Japanese society.
Acknowledgements The author thanks Bob Anderssen in CSIRO, Australia, for stimulating
discussion about mathematics in- and for industry and for his helpful comments on the article.
He also thanks my colleague Yasuhide Fukumoto in IMI for his useful comments on the article.
He finally thanks Seiko Sasaguri for preparations of figures in the text.
References
1. Hosotsubo M, Ito Y, Kuwahara T (2006) Mathematics—the forgotten science. NISTEP Policy

Study No. 12
2. Hosotsubo M (2009) Survey analysis of Japanese mathematics research. J. Math-for-Ind. 1
(JMI2009A)(A-10):73–80
3. OECD/Global Science Forum (2007) Mathematics in industry. http://www.oecd.org/document/
8/0,3746,en 2649 34269 42626653 1 1 1 1,00.html
4. Duff I, Cuminato JA (2011) Brazilian applied math targets local problems. SIAM News
(September) 44–47
5. Wakayama M (2012) Interfacing educational & research with mathematics-for-industry: the
endeavour in Japan. In: Damlamian A, Rodrigues J-F et al (eds) Educational interfaces between
mathematics and industry (EIMI). An ICMI-ICIAM study 20. Springer, Berlin (in press)
6. Lamoreaux SK (1997) Demonstration of the Casimir force in the 0.6 to 6μ m range. Phys Rev
Lett 78:5–8
7. Schuurmans MFH (2010) Casimir and lessons for innovation. In: van Dijk G, Wakayama M
(eds) Casimir force, Casimir operators and the riemann hypothesis. De Gruyter, pp 7–19
8. Rivest RL, Shamir A, Adelman L (1977) A method for obtaining digital signature and public-
key cryptsystems. Technical Memo LCS/TM82, 4 April 1977. MIT Laboratory for Computer
Science (Revised 12 December 1977)
9. Anderssen RS, de Hoog FR (1984) A framework for studying the application of mathematics in
industry. In: Neunzert H (ed) Proceedings of the conference mathematics in industry, October
1983, Oberwolfach, B.G. Teubner, pp 7–34
10. Mumford D, Desolneux A (2010) Pattern theory. AK Peters, Natick
11. Yamamoto M (2013) Mathematics for industry: principle, reality, practice, pp 77–99 (in this
volume)
A Brief Biography of Masato Wakayama, Mathematician
Masato Wakayama received his Ph.D. at Hiroshima University in 1985. His main
research fields are representation theory and number theory. He is a Professor of
Mathematics, Kyushu University since 1997 and now a Distinguished Professor.
He is the founding Director of the Institute of Mathematics for Industry (IMI),
which is the first research hub to bring together mathematics and industrial
technologies in Asia. In addition to the Advanced Mathematics Technology, Applied
Mathematics and Fundamental Mathematics divisions, the Laboratory of Advanced
Software in Mathematics it includes the Office for Promotion of Collaboration and
Consultation, which promotes smooth interaction and valuable joint research with
the industry.
He currently serves as a Vice-President of Kyushu University, and is also
the Program Leader, Global COE Program “Education and Research Hub for
Mathematics-for-Industry.”
Mathematics for Business and
Business Leaders Based on Mathematical
Intelligence
Hiroshi Fujita
Emeritus Professor,
University of Tokyo
Introduction
Japanese business leaders who are authors of the chapters of this book report the
role and contributions of mathematics for their companies or corporations and
relate their experiences of making serious decisions as business heads with the aid
of mathematical thinking. Their comments encourage us to develop mathematics
for business and urge on our efforts to improve mathematics education to foster
potential business leaders.
The present chapter is intended to analyze and respond to their views and
suggestions from the standpoint of mathematics. Except for a few remarks, no
specific mathematical background in readers is assumed below.
We shall begin with recognition of the basic two ways in which mathematics
contributes to business, namely, one as a mathematical tool and the other in
mathematical thinking. For the latter, which is the main target of this chapter, I
propose renaming it mathematical intelligence with two components: a component
of mathematical literacy and a component of mathematical thinking, which will be
described in the following section, Mathematics for Business and Businesspeople.
The remaining part of this chapter is devoted to relevant topics of a somewhat
specific nature, which include subjects such as business mathematics viewed
from the methodology of applied mathematics (in the Methodology of Applied
Mathematics section) and a historical view of great peaks of mathematics as origins
of useful methods (in A Historical View of Great Peaks of Mathematics). Then,
mathematics education to cultivate mathematical intelligence will be discussed
124 H. Fujita
in regard to high school education and liberal arts education (in Education to
Foster Mathematical Intelligence). The chapter concludes with some comments on
the Japanese cultural background concerning mathematics as compared with the
Western one.
Finally, I am afraid that, because of my specialty and career, the scope and
content of this chapter is rather biased to applied analysis, if it should speak for
mathematics in general.
Mathematics for Business and Business People
Consideration of “mathematics for business” as a whole would require specific

discussions on various areas including accounting, econometrics, and financial
engineering, and also general views of economics and management. Because I am
not qualified to carry these out, I shall be concerned only with those topics that have
been taken up by business leaders in this book.
According to almost all the chapters, the ways in which mathematics contributes
to business belong to the following two categories: (1) as mathematical methods and
tools, and (2) through the means of mathematical thinking. Namely, the contribution
to business comes from the two facets of mathematics, “Ars” (techniques; technical
methods) and “Logos” (language; intelligence).
To put it concisely, in the first category, a company or corporation benefits
from using standard or advanced mathematical tools, while in the second, its
members, particularly the leading members, exert thought processes mathematically
to understand problems, to make decisions, and to communicate with their fellows.
It should be noted that effects of these two categories are not mutually exclusive;
they usually are intertwined with each other in real situations.
For example, while Toyota’s “Kaizen” activities are concerned with practical
tools and processes, those tools and processes can neither be utilized effectively
nor improved without mathematical thinking. Likewise, a mathematical way of
thinking, which is called for in companies’ management decision making, can be
more effective if it is aided by mathematical tools such as computer-driven statistical
processing, visualization, and simulation.
There are a number of mathematical tools that are already standard and are
utilized in business on a daily basis. It is interesting, from the standpoint of applied
mathematics, that some tools are based on recently developed mathematics whereas
others are uniquely revived applications of classical principles of mathematics (e.g.,
see the chapters by Hasegawa, Wada, and Kawano).
In addition to conventional purposes of mathematical tools such as product
improvement and process streamlining in manufacturing, mathematical tools are
recently being used with increased importance in order to meet new difficulties such
as ensuring safety and reliability and dealing with risks caused by human, natural,
or social factors (e.g., see the chapters by Watanabe, Wada, Masada).
Mathematics for Business and Business Leaders Based on Mathematical Intelligence 125
The following two quotations from chapters in this book are lively examples of
ideal collaboration between mathematical tools and creative minds of leaders, which
is realized in typical venture companies.
When describing natural phenomena, we human beings have been using mathematical
expressions to simplify the principle, not only in physics and chemistry but also in biology,
thereby expressing phenomena intuitively and in an objective fashion. Here, “to ascertain
the essence of things” and “to identify the true nature of things” which underlie Kyoto
culture seem to coincide amazingly well with methods of mathematical expression.
(Horiba)
The goal of our company is to actualize artificial intelligence. Inventing A.I. technology
means to research the function of the brain while describing the finding in mathematical
language. (Iwane)
In the next section I shall discuss the second category, i.e., the way of mathemat-
ical thinking, in more detail.
Mathematical Intelligence Composed of Mathematical

Thinking and Mathematical Literacy
By a closer look at the above-mentioned business leaders’ discussions and through

comparison with the nature of mathematical sciences, i.e., wide-ranging applied
mathematics, it is revealed that the ability of so-called mathematical thinking
consists of two elements, namely (1) a thinking ability, which is parallel to a proper
mathematical way of thinking, to formulate a problem mathematically and solve it
by means of mathematical logic and (2) an ability of logical language to make use of
mathematical concepts and expressions for the purpose of understanding a problem
quantitatively, giving objective explanations, and increasing the persuasiveness of
management judgments.
I call the first element the ability of mathematical thinking and the second
element mathematical literacy [2–4]. Incidentally, the ability to understand and
make use of numbers and numerical variables, often called quantitative literacy,
should be considered as part of what we call mathematical literacy. Furthermore, the
ability to understand and use simple numbers, which is called numeracy, is where
mathematical literacy starts to develop in one’s childhood.
After these preparations I would like to define mathematical intelligence as
a combined ability of mathematical thinking and mathematical literacy. Figure 1
shows relations among these concepts. The “Logos” of mathematics contributes to
business through exertion of mathematical intelligence, as is pointed out by Fujio
Cho, the chairman of Toyota, as below.
Without numbers, we can say nothing but subjective remarks like “That vehicle looks
cool” or “This one provides a comfortable ride.” When we have to develop a new car,
we cannot depend on such empty words. What we need is information that is expressed
126 H. Fujita
Fig. 1 Two foci of mathematical intelligence
with numbers, such as bulk weight, noise level, vibration level, etc. By “numbers,” I am
not talking about technical calculations that are conducted in laboratories. I am talking
about simple arithmetic, such as addition, subtraction, multiplication and division. Numbers
and arithmetic are what Toyota’s 300 thousand employees share as the fundamental
communication tool. (Toyota)
Mathematical Thinking for Business
While mathematical thinking should play its role in every phase of business and in
every corner of companies, its most serious and challenging stage is cases where
business leaders need to make the best decision in a complicated situation, taking
future risk factors into account and foreseeing the big picture, within a limited time,
as many of the authors of this book who lead their businesses would agree.
Needless to say, aside from some expectation that activities based on mathemat-
ics for business may stimulate the interest of mathematicians as new targets, the goal
of mathematical thinking in business is not aimed at the progress of mathematics
itself. Nonetheless, the ability of mathematical thinking that business leaders are
required to exercise is comparable, in its style and depth, to that of professional
mathematicians in research. In fact, both mathematicians and businesspeople are
dedicated to finding the essence of problems and thinking through them until they
reach a significant solution.
The capacity enabling such efforts is so basic in nature that it can be developed
initially by eager learning of high school mathematics. This point is well stated by
Norio Wada, the chairman of NTT, as below.
By learning school mathematics using numbers and figures, we have acquired these thinking
processes, i.e., the ability to look at objects from a variety of perspectives, and ultimately
discover a truth that suits the intended purpose and express it in a simple way. Once we
graduate from school, we are quite often confronted with problems that we have not studied
in school, and become highly upset due to the fact that we don’t have a clue as to what to
do. My experiences show that the mathematical thinking mentioned above is very helpful
in solving problems in such situations. In other words, mathematics is a tool that enables
us to broaden our ability to deal with problems that confront us, and makes our conclusions
highly convincing. . . . I’d like to emphasize the importance of mathematics education in
laying a foundation for teaching people to look at and think about things from perspectives
not confined to any existing framework. (NTT)
The appreciation of mathematical thinking expressed by business leaders seems

to be supported also by their feeling to respect courage and honesty, which are
needed when one recognizes and accepts the pros and cons of one’s own analysis.
Such an attitude of business leaders shares much with scientific morality among
mathematicians, which encourages me again to describe their way of thinking as
mathematical.
Here let me confirm that the business world appreciates the mathematical way
of thinking for realistic problems, and then proceed to itemization of typical cases
where mathematical thinking is applied.
1. Meet and Face the Problem.

Business leaders are not like students who solve problems that are prepared by
somebody else. As the authors of this book point out, problems in the business
world cannot be met in a routine or straightforward manner. Rather, business
practitioners are forced to face uncooked problems that consist of vast information,
complex phenomena, and tangled conditions. First of all, they must find the
problems’ essences with an objective mind and a method of abstraction, which
are characteristic in mathematics. This should be carried out with an attitude
to avoid stereotypes, by introducing necessary concepts and having a broad and
flexible vision.
As it happens in mathematics, a creative breakthrough against tough problems
can be made here also by going back to the origin of the problems and reformulating
them all over again. Namely, go back and ask again! This recommended practice
is similar in spirit to Toyota’s Kaizen slogan “Ask ‘why’ five times,” which is
mentioned in the chapter by Chairman Fujio Cho in this book.
2. Think over the Problem and Solve It.

The moral of mathematics in problem solving is to think about it over and over again
with patience until getting a flash of inspiration. This kind of flash would be fired
often by becoming familiar with the works of eminent predecessors and updating
their original ideas in order to suit them to present cases, as well as by exercising
keen insight to build up a general overview through careful investigation of special
cases.
128 H. Fujita
The business leaders in this book confess that the difficulty of finding a solution
in business under given conditions is reminiscent of “equation solving” that they
experienced in their high schools. Let me quote Yasuchika Hasegawa of Takeda
Pharmaceutical Company as an example.
In managing day-to-day operations, there are often times when I get the feeling that I
am subconsciously engaged in solving a system of equations. These can be multivariate
equations that have a significantly higher number of variable parameters than those given
in the above equation, equations with many thousands of constants and variables. There are
times when it is literally impossible to pin all the variables down. (Takeda)
Fostering of Mathematical Thinking Ability
How can the ability of mathematical thinking as required in business be developed

and strengthened? While individual effort after graduation would certainly be of
value, the basic ability of mathematical thinking is generally fostered by mathe-
matical education at high schools and universities. Thus mathematical education
assumes a heavy responsibility. Among others, problem-solving exercises in high
schools and liberal arts education oriented toward mathematical literacy during early
years at universities would be especially important.
With their flexible mindset, good high school students are able to acquire the
basics of mathematical thinking through early exposure to the scientific essence of
mathematics. Even at the level of high school mathematics, their heuristic intuitions
can improve dramatically if they do not run from difficulties and think over tough
problems with patience. A recollection of Fujio Cho, as quoted below, seems to
reinforce this belief of mine.
When I was in high school, “doing mathematics” and “being logical” seem to me to be rather
the same thing. Doing mathematics made me develop a habit of seeing things logically.
When studying math, we have to think, not memorize. . . . As a high school student, I was
trying to solve trigonometry problems. They were very difficult ones. I had been thinking
about one of these problems for a whole day and finally found the answer. It seemed
something inside me had changed after taking a whole day to solve the first problem.
(Toyota)
I will also quote from Koichi Hori’s (Dream Incubator Inc.) essay “Mathematics
Is Indispensable to Business Managers,” which appeared in the Japanese edition of
this book. He makes a proposal on the basis of his experience of having studied at
The University of Tokyo and Harvard Business School and succeeded as a business
consultant.
In our time, when we can so easily benefit from the progress of computer technology as
well as mathematical tools such as accounting and business science, we would give up
any privileges of the era if we steered the business without recognizing the importance of
mathematical thinking. . . . I dare say that universities only have to provide science courses
(in the broadest sense of the term, which includes engineering) to undergraduates. Study
of specialties such as law and economics should start in graduate school (master courses).
Undergraduate curricula should be organized with their cores in science and with emphasis
on mathematics and logic, and further on learning the scientific methodology of natural
science to set up a hypothesis and test it. (Dream Incubator)
Most of the business leaders who appear in this book spent their high school
and university years in the 1960s and 1970s; it was the era when mathematical
education and the social environment in Japan were in good shape. Students at
that time used to read more books and study harder than today. With remnants of
the pre-World War II educational framework still in evidence in schools, teachers,
and parents, high expectation for learning was shared among students. Besides, the
open atmosphere after the educational reform following the war had made it easier
for excellent students to hold big dreams to realize diverse possibilities. Since the
ability of critical thinking was highly weighted in entrance exams of top universities,
which already were seriously competitive then, high school students were required
to have an academic attitude even in their exam studies. It is unfortunate that the
current condition of Japan’s mathematical education and social environment are
deteriorating noticeably.
Mathematical Literacy for Business
Mathematical literacy is, just as the term indicates, the ability to utilize basic
mathematics linguistically. The word is now widely known as one of the survey
items in the Program for International Student Assessment (PISA). I was the first
advocate of mathematical literacy who introduced the word in the mid-1980s [2, 3]
and maintained that it should be a part of the attainment target of secondary school
mathematics.
I have stressed that the principal goal of mathematical education in high schools
is to help students develop their mathematical intelligence, which is, in parallel to
discussions made above (in Mathematical Intelligence Composed of Mathematical
Thinking and Mathematical Literacy), composed of mathematical thinking ability
and mathematical literacy. In this section, I will verify the importance of mathemat-
ical literacy for businesspeople.
As stated earlier, by mathematical literacy for businesspeople, I mean the ability
to leverage mathematical concepts and expressions to nail down a problem in a
quantitative way, to ensure an objective explanation, and to improve the persua-
siveness of business judgments. The merit of mathematical exposition, including
numerical expressions, is certainly its clarity and objectivity. It helps us differ-
entiate true and false unambiguously. Being free from subjectivity, cheating, and
fabrication, mathematically proved statements are strongly persuasive to anyone.
They are not only useful in inter-organizational and interdisciplinary cooperation,
but also provide a service in international communications where people do not
necessarily share the same cultural background and the same language. This is one
130 H. Fujita
of the reasons why the business leaders in this book highly regard mathematical
literacy, as exemplified by the following quotation.
No matter how talented or experienced, no company president should make such decisions
in isolation. The organization of a business thus allows a company to allocate roles and
responsibilities and to delegate authority. Through this system and the integration of
specialist knowledge and skills, a company is best able to handle the issues it faces. . . .
It would be wrong to suggest that executives and managers simply embrace the qualitative
intuition derived from experience in reaching key decisions. It is here that the enormous
contributions are made by quantitative factors. Here, mathematical modeling, which is the
subject of this discussion, comes into play. (Takeda)
Traditionally, scientists have recognized mathematics as the common language

for science. As is well known, Galileo Galilei said, “The laws of Nature are written
in the language of mathematics.”
Also the Japanese Nobel laureate Shin’ichiro Tomonaga defined physics as
“Study to discover natural laws through experiments and observations on natural
phenomena, and to describe them in terms of mathematics.”
At the same time, business leaders point out that the importance of mathematics
lies also in its usefulness as a basic tool to facilitate international and/or interdis-
ciplinary cooperation, which sounds fresh and novel to me. To illustrate this point
of view I would like to quote from the chapter by Ken’ichi Horiba, who leads the
Horiba Group, whose non-Japanese employees comprise more than 50% of the total
employees.
The technical network in our company is formed on the basis of communications. . . .
I enrolled as a junior in the Department of Electrical Engineering in the College of
Engineering at the University of California Irvine (UCI). At the time, I had a very hard
time coping with the English language. However, the mathematics I studied at the university
in Japan was helpful. There are no language barriers in mathematics. . . . The products of
our company are based on science including mathematical principles. Our executives and
our managers as well as many employees have scientific backgrounds. . . . Engineers and
top management overcome these differences between regions while being connected on the
basis of mathematical principles and one corporate culture. (Horiba)
I wonder if the reason that Japanese business leaders emphasize the importance
of mathematical logic so strongly might originate from the fact that the Japanese lan-
guage is more implicit and feeling-based than explicit and logical when compared
with European languages. Even a Japanese mathematician can sometimes explain
his research more clearly in English than in Japanese. Kaoru Yosano, the only
politician who contributes to this book, writes about how English and mathematics
can be harmonized so easily.
When I was an eighth grader, my father, who was a diplomat, was dispatched to Egypt, and
I was put into a British boarding school in suburban Cairo. From the ninth grade, I started
taking classes in English. Since the school rules were strict and I could not understand
English well, I had a hard time of it. In those days, however, I made the wonderful discovery
that math was easy when taught in English. To solve a word problem in math, for example,
you have to ask yourself questions such as “what are the given conditions?” and “what
is the unknown value?” I could understand these processes more clearly than when I was
studying math in Japanese, because both math and English were languages that emphasized
the importance of clarity. After a short time, I became the best math student in my grade.
(Yosano)
Finally, I would like to remark that so-called mathematical literacy has a

certain nuance in the ability of a common user to apply mathematics, which thus
resembles the meaning conveyed by the term “computer literacy.” At the time I
advocated the importance of mathematical literacy as one of the main goals of
high school mathematics, the emphasis was placed on acquiring a mathematical
background that is adequate to use mathematics in one’s daily life at home and in
the workplace. Similarly, with the term mathematical literacy for businesspeople, I
hope intelligent businesspeople will have the abilities to use the mathematical way
of communication and fundamental problem solving in their areas of business, even
though they are not professionals in mathematical sciences and are of the level of
concerned laypersons with regard to mathematical knowledge.
Methodology of Applied Mathematics
Three Keywords to Review Methodology
Traditionally, mathematics has been applied to understand phenomena of the outer

world, especially nature, as well as to fulfill the needs of human society. A typical
example of the former case would be mathematical methods developed in basic
science such as physics. A time-honored example for the latter is the mathematical
way of problem solving in technological fields, such as treatment of dynamical
systems and distributed parameter systems in engineering.
Integration of science and technology is advancing today through deepening
scientific treatments and developing methodologies in diverse areas, which was
boosted by utilization of computers. A trend toward “unified science” (an integrated
form of science and technology) is observed in various areas in an interdis-
ciplinary way. Being part of such unified science, mathematics is expected to
make contributions to it in addition to just developing itself. In Japan, the unified
field of mathematics formed by expansion of research areas, which includes pure
mathematics as a core, with application-oriented intentions, is called mathematical
science. Namely, mathematical sciences stand for wide-ranging, comprehensive,
and interdisciplinary applied mathematics to respond to missions of our time. Thus,
mathematics for business, which is the topic of this book, is identified as a newly
developing field of mathematical sciences in its methodology.
Note: In Japanese, when emphasis is placed on practical applications especially

in engineering, we tend to use the term “mathematical science for application”
instead of “mathematical sciences” or “applied mathematics,” the latter being apt
to imply technical application of existing mathematics. For example, the Japanese
132 H. Fujita
counterpart of SIAM (Japan Society for Industrial and Applied Mathematics)

has a Japanese name whose direct translation into English is “Japan Society of
Mathematical Science for Application.”
Here, I would like to refer to the article “Mathematics Contributing to Innovation

of Management” written by Shiro Hiruta of Asahikasei in the Japanese edition of
this book about the mathematical methods of solving problems in the manufacturing
industry:
Not only the mathematical way of thinking, but also simulations assisted by mathematical
methods, is quite effective in solving problems. The latter is utilized in various fields,
including detection of causes of troubles, optimization of expected performances, and best
possible adjustments of usage conditions. Conversely, without the aid of mathematical
methods, our problem-solving effort will get stuck most probably . . .. (Asahikasei)
From the point of view of methodology in applied mathematics in the sense of

mathematical sciences (henceforth “applied mathematics” for short), we notice that
prior to the application stage of mathematical methods, there is a stage in which
we view the problem to be solved or the phenomenon to be clarified and grasp
it as a problem of mathematics. It may be also called the stage of mathematical
modeling. Furthermore, it coincides with the stage of “expressing a phenomenon
mathematically” according to the words of Mr. Kazuo Iwane introduced in the
section above, Mathematics for Business and Businesspeople.
Now, I would like to choose the following three keywords, which are useful
in our consideration of overall activities (research and/or development) of applied
mathematics [4, 5]:
Object, Concept, and Method.
A few words of explanation are in order. “Object” is either a problem to be solved
or a phenomenon to be clarified. In research and development in cooperation with
other fields, “object” is not limited to those in mathematics. The reason I dare to
use the term “object” is because I wish to cherish the feeling of reality, the reality
including the background of a phenomenon or the origin of a problem.
“Concept” is a linguistic component needed to formulate a phenomenon or a
problem under consideration, namely, to model a problem mathematically. While
many of the concepts originated in mathematics are based on definitions, those
in interdisciplinary research and development include jargon imported from the
terminology of other fields.
Lastly, “method” stands for a way or a technique to analyze or solve a
mathematical model of the phenomenon or problem under consideration.
As an example of use of these keywords, let us look at the clarification of the
movement of planets by Newton, which is the first great success of human beings in
mathematical sciences.
Here, “object” is the clarification of the movement of planets around the Sun.
The mathematical model used in this case is made up of those differential equations
that specify the movement of particles under the influence of universal gravitation.
The “concept” that enabled such modeling is, on the mathematical side, differential
calculus (invented by Newton himself) and acceleration, etc.; and on the physics
side, particles and universal gravitation (the law discovered by Newton himself).
The “method” used by Newton to solve the problem was differential and integral
calculus by hand.
Next, as an example from advanced technology, I would like to mention the
chapter “Improvement of Railway System Technology” by Mr. Eisuke Masada,
chairman of the Railway Technical Research Institute, in this book. In this case,
“object” is the task to realize a high-performance train including a linear motor
train. To build up mathematical models that help them realize the expected high
performance, they use concepts from physics and engineering. The conditions
posed in order to formulate the task are equations of motion in complex dynamical
systems and partial differential equations such as electromagnetic equations. As for
“method” solving them, they use the theory of partial differential equations and
computer simulations.
Methodology of Applied Mathematics
Although the situation has been exemplified through the preceding examples, here
I would like to analyze the methodology of mathematical sciences, or modern
applied mathematics from a general point of view [4, 5], and in accordance with
the main subject of this book. In doing so, I hope that the here-stated arguments on
methodology of applied mathematics would be suggestive also in examining the use
of mathematics for business and in confirming the way of mathematical thinking by
business leaders.
Actually, when we encounter phenomena to study in the light of mathematics
or problems to solve with the aid of mathematics concerning certain objects, we
start with mathematical modeling of the latter. Namely, modeling is the first stage
of research (and development).
There, proper concepts must be called for, and we formulate the problem into a
mathematical model. In carrying out this process, it can happen that some concepts
are taken from outside of mathematics and that the model is formed with the aid
of laws or formulas in related fields of science. In these cases, a good model
is an outcome of the researcher’s multidisciplinary scholarship or his/her close
cooperation with fellow experts in relevant fields.
Next to modeling, which formulates the object mathematically, the second stage
comes up. Namely, the mathematical model thus obtained is analyzed or solved
by means of mathematical methods that are chosen by the researcher, often in
consideration of computing cost and time. The necessary mathematical work in
order to calculate or to solve something is carried out, very often in the form of
computer simulation. Thus this stage of mathematical analysis is sometimes called
the simulation stage.
134 H. Fujita
The choice of suitable methods depends on researchers’ scholarship, experience,

and wisdom. We hear from engineers that in dealing with practical problems for
industrial applications, a clever combination of (mathematically) simple methods
can yield a good result very often.
This means that good engineers (and good businesspeople) should have a
mastery of mathematical literacy concerning their possible tools. On the other hand,
particularly, analysis of a new type of model is a good chance to obtain a new
method by improving existing methods or by inventing an original new method, both
through hard thinking. Mathematical thinking ability is called for in these cases also.
In this way, analysis by the method is accomplished. Then we proceed to the
stage of conclusion or feedback. Namely, the outcome of mathematical analysis is
examined. Then the following cases would come up. First, the case in which the
result is a satisfactory success. The second case is such that the result is significant
but is insufficient. The third case is one where the study or project turns out to be a
failure and must be abandoned.
In the second case, which is most often realized, we feed the outcome back to the
proceeding stages. Namely, feedback is done in order to improve the model or the
method.
Thus the progress is sometimes advancing and sometimes iterative.
The following diagram shows the flow of work along the stages of the method-
ology. Namely,
modeling (start) −→ mathematical analysis −→ conclusion or feedback
Some remarks on the above-mentioned methodology of applied mathematics are

in order.
Remark 1 (Similar Methodologies in Various Fields). Even though the term of
each stage differs from one field to another, this type of methodology is effective
in research and development of various fields, ranging from basic science to
practical technology. For example, in experimental physics the initial stage of
“setting a problem and designing an experiment” corresponds to modeling. Further,
implementation of experiment and measurement of concerned data corresponds
to carrying out the stage of mathematical analysis in applied mathematics, to the
effect that it is an effort to obtain the outcome of the method. Moreover, the
branching of the process of the experiment to termination or the continuation with
a modified setting corresponds to the branching stage that separates into conclusion
or feedback.
Incidentally, in theoretical physics, the modeling stage is usually called formula-
tion. This could be due to the fact that in many cases the model is uniquely specified
by fundamental equations in physics.
Contrastively, in fields of applications (including business mathematics), it is
often up to the mindset of a researcher to decide how to extract a problem from
complicated phenomena and conditions; certainly there the term “modeling” seems
to be more appropriate.
Remark 2 (Prototype in Elementary Mathematics). Let us look at word problems
in high school mathematics. Then, understanding the point of a problem and formu-
lating equation(s) correspond to the modeling stage, while solving the equation(s)
thus formulated is the stage of mathematical analysis. Correcting the formulation
and improving the calculation method after checking outcomes, correspond to
the feedback stage. This similarity seems to highlight the naturalness of the
above-mentioned methodology in applied mathematics. Furthermore, it would be
instructive if we make high school students understand that word problems carry
such a prototype of scientific meaning for future development.
Remark 3 (Methodology in General Natural Science). “Make a hypothesis and
verify (test) it,” which is said to be general methodology in natural science. In
fact, it is methodology for scientific clarification of the essence of a phenomenon
rather than methodology for solving a problem. In this sense, the mathematics
counterpart to “making a hypothesis” is to conjecture an interesting theorem. And
the mathematics counterpart to “verifying a hypothesis” is to prove a conjectured
theorem. When it comes to simulation approaches to understand phenomena
mathematically, “make a hypothesis” and “verify” are to be replaced by “construct
a model” and “test the validity of the model through simulation,” respectively.
Remark 4 (Positive Meaning of Feedback). Feedback in the above-mentioned
methodology is not limited to just a remedial step, such as “correction of deficiency.”
An act of modifying a model for obtaining a desirable solution has a positive
meaning that goes hand in hand with the spirit of control theory, where parameters
are modified recursively depending on outcomes of the tests, in order to obtain
136 H. Fujita
more satisfactory solutions of the problem. Indeed, J.L. Lions, who was a leading
scientist of applied mathematics, referred to the third stage of the above-mentioned
methodology as feedback and control. The positive meaning of feedback seems to
have great significance for problems of time evolution systems, like the prediction
problems about investment and management in business mathematics, where one
needs to calculate the solution of a problem at every instant along the flow of time.
Characteristics of Mathematics for Business Seen

from Methodology
From the point of view of methodology of modern mathematical sciences argued in

the two previous subsections, we are going to compare mathematics for business,
which is destined to be mathematics to meet real problems of the business world,
with traditional applied mathematics for science and engineering. Then it will
be revealed that there are remarkable differences. Let us point out several of
them here.
1. Mathematics for business often deals with “varying objects.” Especially, changes
accompanied by future risks pose a serious object for consideration. This is a
dynamical problem in its nature and cannot be fully covered by static methods
such as usual linear programming. Moreover, the concept of “risk,” in the first
place, has not shown up as a target of analysis in standard mathematics and
physics even when it was concerned with time-evolving phenomena.
2. In the analysis of mathematics for business, which traces the “change” in time
evolution, the solution of a problem is not the only element that changes in
time. Conditions defining the problem may change, which is due to various
factors that are difficult to predict in advance, such as changes in society and the
market, or changes in strategies of competitors. Analysis incorporating changes
in the outer environment of business in real time is necessary. If we can put
it mathematically with the terminology of differential equations, it means we
are required to set time-evolution equations where “coefficients” and “boundary
conditions” vary with time. Moreover, future coefficients of the equations may
be affected by the behavior of the present and past solutions. Thus one faces
the need for taking into account the feedback from the knowledge about past
solutions to reformulate the setting of the problem. Consequently, any expression
of these (time) evolution systems in the form of differential equations will
inevitably exhibit nonlinearity. Moreover, some integral-equation-type models
accompanied by nonlocal conditions on the solution will be necessitated.
3. There are various kinds of “risk” as well. While there are predictable risks such
as economic loss caused by normal changes, there are quite huge and singular
risks due to natural disasters, man-made accidents, or structural instability; the
examples are great earthquakes, giant tsunamis, nuclear power plant accidents,
and financial crises. Speaking of a mathematical model, adoption of equations
with singular solutions that behave somehow catastrophically is required. The

research of these types of mathematical models includes the study of “blowing-
up of solutions” which was initiated by the present author [1] and Prof. Y. Giga
[6], who is one of the editors of this book.
In mathematics for business, a real-time response is required even to the
kinds of risks such as those as mentioned above. Regarding this point, efforts
to adapt the control theory in engineering, which has renowned achievements in
controlling the movement of rockets, to the desired goals in our context, would
be necessary and effective.
4. Companies aim at maximizing profit while suppressing cost. Consequently,
a traditional problem in mathematics for business is the task of seeking the
maximum under given conditions. It is the effort to attain an optimal solution
under the imposed realistic constraints. On the other hand, in mathematics, the
search for the maximum/minimum is a classic problem. There is a long history
of mathematics flowing from a maximum/minimum problem of a function which
is familiar even to high school students, to the calculus of variations, which is an
important topic of modern analysis. In actual problems to find an optimal solution
in mathematics for business, the number of variables and conditions is quite
large. Although mathematically the existence condition of optimal solution is a
concern, it is more urgent in practice to find and establish an efficient algorithm.
This task is an important and challenging topic in applied mathematics.
5. Lastly, problem solving in mathematics for business requires promptness of
action as mentioned above. As for research by mathematicians, essentially no
limit is imposed on accomplishment (the academic dissertation by a graduate
student has a deadline for submission, but it is only for premature mathematicians
like them). In the research and development of engineering, the progress schedule
chart of the enterprise will be taken seriously. In contrast to these, in practical
situations of mathematics for business, appropriate actions must be taken on time
in consideration of any information available. This might be said to resemble
military strategy since the old days; intuition and wisdom of the executor that
transcend logical thinking must play a crucial role. This aspect of the matter
calls for mathematical intelligence of the executor over simple awareness of the
methodology of applied mathematics.
A Historical View of Great Peaks of Mathematics
Three Peaks in the History of Mathematics
Many of the mathematical concepts, methods, and ways of thinking that support
the application of mathematics are the outcomes of outstanding great works of
mathematics, so to speak, the great peaks in history of mathematics from ancient
138 H. Fujita
times. Reviewing these great peaks would be useful and inspiring for us to found
reliable and progressive mathematics for business.
With an intention to project the result on our efforts to develop business
mathematics in the 21st century, and to encourage mathematics education to
cultivate mathematical intelligence in students who are potential business leaders,
we shall here examine such alpine scenery of great achievements of mathematics
and recall characteristics of each of these peaks briefly [4].
It is well known that ancient mathematicians in Arabic counties, China, India,
and other parts of the world made great contributions. We could, however, count
the following three as giant peaks in the history of mathematics, as far as direct
influences on modern mathematics and its application are concerned.
1st peak: Birth of Euclidean Geometry in about 300 BC.
2nd peak: Discovery and Development of Calculus (Differential and Integral
Calculus) in the 17th and 18th centuries.
3rd peak: Development of Modern Axiomatic Mathematics in the twentieth
century.
Incidentally, recognition of the importance of these three peaks in regard to
mathematical education was first proposed by the late Professor Yukiyoshi Kawada
around 1970, when the wave of the New Math movement reached Japan. The
suggestion of Prof. Kawada was seriously considered in organizing the Japanese
national curriculum of high school mathematics for those periods, when most of the
business leaders who are authors of this book learned mathematics as high school
students.
Characteristics of the Three Peaks
(a) Euclidean Geometry

Greek geometry, which handles figures, is by no means just an accumulation of
knowledge about figures; it is known as a model of study because it starts from
seemingly universally accepted axioms and constructs a systematic theory with the
aid of rigorous proofs. A closer look reveals that an idealization of concepts has
been achieved before the formation of axioms; for instance, a point without size
or a line that extends to infinity is an idealization of what we see in the real world.
Prof. Kazuhiko Nishijima, who was a leading particle physicist in Japan, once stated
[8]: “Science is made possible by idealization of concepts and systematization of
knowledge.” Euclidean geometry is a model of study in this respect, too.
While systematization of knowledge in most areas of natural science is guided by
confirmed causality, the systematic structure of mathematics has been formed oth-
erwise. Namely, it begins with taking axioms as their starting point, then introduces
necessary concepts through “definitions,” and finally integrates important properties
of figures systematically in the form of “theorems.” Although the theorems are
established by logically strict “proofs,” the importance of each theorem is subject

to the value judgment of mathematicians. In this respect, Euclidean geometry is a
model of modern mathematics.
Actual work for systematization in mathematics calls for inductive and heuristic
thinking. An idea for a significant theorem initially comes up through a deep insight
into characteristic properties of the objects under consideration. Giving a proof to a
conjectured theorem often necessitates a discovery of a creative method, a prototype
of which is “discovery of auxiliary lines” in Euclidean geometry.
Incidentally, I note that what lays the ground of science and makes it so
trustworthy is the recognition that mathematical truth established by rigorous proofs
does not permit any doubt or counterargument as long as the fundamental axioms
are agreed upon. In contrast, if you think of a trial you would notice that claims in
the real world are usually vindicated by evidence instead of a proof.
As early as at the level of the former half of secondary education, students can
touch the prototype of science by studying mathematics such as geometry and
smell the scent of science. This experience of learning is the starting point for
the development of mathematical intelligence, namely, mathematical language and
thinking ability, which should be put into practice when those students grow to
adulthood.
Remark. A traditional area of geometry that describes figures by means of coordi-

nates and equations, and thus examines their property, is called analytical geometry
and has played the role of connecting Greek mathematics and modern mathematics,
as is symbolized by the term “Descartes coordinates.”
(b) Calculus
The word “calculus” as used here means infinitesimal calculus and is used to refer
jointly to differential and integral calculus. In addition, mathematical fields built
on differential and integral calculus are called analysis. The theory of differential
equations and the theory of complex functions are primary branches of analysis. We
shall use here the terms “analysis” and “analytic” in this sense.
The objective of analysis is to study change (variation): namely, change of
numerical variables. “Movement of a point,” which Newton studied at the beginning,
is nothing but the change in the position of a point in the passage of time (an
independent variable). As we have learned in high school mathematics, a change
of values of an independent variable gives rise to a corresponding change of the
dependent variable. According to modern terminology, this correspondence between
the two variables is itself called a function. As for the function describing the
movement of a point, coordinates of a point are dependent variables. Analysis is
a field of mathematics about functions.
Changes include local changes and global changes. When the variable is time,
i.e., when the object under consideration is a system evolving in time, what matters
locally is the changes in the vicinity of a given instant. A typical example is the rate
of change in that instant. If the function represents coordinates of a moving point,
140 H. Fujita
its instantaneous rate of change is none other than velocity. In contrast, in global
changes, we ask the amount of variation of a function when variables vary over a
specified finite or infinite region.
In problems encountered in mathematics for business, we ask the variation of
a quantity in time, make predictions about the future, and try to cope with future
risk factors. Such kinds of problems are those of a time evolution system, and
the treatment will be essentially analytic. In the problem of coding of pictures, as
introduced in this book by Mr. Wada, the president of NTT, they handle variations
in the shades of colors of a picture. It is quite natural that they are using analytic
methods to handle such spatially extended variations. One of the interesting methods
that is realized in this imaging technology is to consider a small increment Δ F of a
function over a small duration of time Δt in tracking the change of an image along a
consecutive frame, a method that is reminiscent of the idea of differential calculus.
In addition, as we know from high school mathematics, one of the central
problems around functions is to find the maximum/minimum of a function. The
extension of the latter, namely, the problems of maximization and optimization, are
often encountered in mathematics for business. They are essentially analytic as well.
(c) Axiomatic Mathematics

Axiomatic mathematics, which flourished in the twentieth century, has influenced
almost all fields of mathematics. Traditional algebra, geometry, and analysis fields
are treated in unity, with the attitude of setting axioms first and then deriving results
as generally and rigorously as possible by means of abstract concepts and methods.
For example, in functional analysis (an abstract version of analysis), a function is
considered as a point in a set called functional space; if the functional space is a type
called Hilbert space, the concept of orthogonality of two functions is introduced,
and a generalization of Pythagoras’ theorem plays a fundamental role. Operators
concerning differential equations and integral equations enter the stage there as well,
and manipulation of operators is conducted algebraically.
When compared with the features of the above-mentioned Euclidean geometry,
modern axiomatic abstract mathematics is, in its spirit, a descendant of Greek
geometry. Although the freedom in choosing objects of interest on the basis of
appropriate axioms and the attitude of searching deeply into the essence can be
regarded as a virtue of modern mathematics that we can be proud of, it also shows a
tendency for ideas to turn too far to inward speculations, and connections to natural
phenomena and real society get weaker.
Mathematical Sciences, a Great New Alpine Plateau
Coming back to reflections on the historical scale, and as a main point of this section,
we claim that a new peak has shown up recently. It is the computer-driven advance
of mathematical sciences. In its wide range and its potential height, the new peak
deserves the title of the fourth giant peak in the history of mathematics, although it
might be better to use the term “a great alpine plateau” instead of “a giant peak.”
Thus we regard the following four as the towering high-rises of the mathematics
concerned:
Euclidean Geometry, Calculus, Axiomatic Mathematics, Mathematical Sciences.
In this direction, we believe, mathematics will find its promising way with its
own vivid creativity and with recovered links between mathematics and the real
world. Formation of advanced and trustworthy mathematics for business should be
strived for in this spirit.
At this point, however, we note that as to mathematics, the rise of a big new
peak does not mean that the former peaks are wiped out, which the very history
of mathematics shows. As for mathematics, such a big change takes place, so to
say, as sublation (Aufheben) rather than a drastic paradigm shift observed in some
fields. The heritages of modern mathematics remain important and will be respected
even in the era of mathematical sciences, although stagnation caused by excess
ramification inside and by unconcerned isolation outside must be cured.
In this connection let us quote a passage by Poincaré [9].
If I may be allowed to continue my comparison with the fine arts, the pure mathematician
who should forget the existence of the exterior world would be like a painter who knew
how to harmoniously combine colors and forms, but who lacked models. His creative power
would soon be exhausted.
The second point we want to emphasize here is that this new plateau, the fourth
peak, shares basic features with the second one, the peak of calculus, in the sense
that both are characterized by (1) realistic concepts, (2) powerful methods for rich
and far-reaching applications, and (3) tight linkages with outer fields and objects. In
short, the fourth peak of unified mathematical sciences is outgoing in its character
just as the second peak of calculus was.
Education to Foster Mathematical Intelligence
There is no doubt that mathematical education in high schools and universities plays
an important role in training students, who would become experts in science or
engineering in industry, in order to enable them to acquire mathematical capability
at the required level. On the other hand, I have already discussed briefly in the
second section above, Mathematics for Business and Businesspeople, the role of
school education to foster mathematical thinking ability and mathematical literacy
of relevance to general business leaders. The present section is a supplement to my
earlier argument there.
142 H. Fujita
Mathematical Thinking Ability and Math-Sense
At the educational stage, “understanding mathematics” is not equivalent to having

just plenty of mathematical knowledge nor to learning many patterns of problem
solving by heart. Rather, it should imply that one is able to think out mathematics
and get it by his/her own mind. The reason I have to make such an obvious confir-
mation is because an inclination toward “memorizing mathematics = mathematics
without thinking” recently has spread among Japanese students, which results in a
serious decline in the level of mathematics education in Japan. It was Prof. Kunihiko
Kodaira, the first Japanese mathematician who received the Fields Prize, who issued
a warning on this tendency as early as 1984 [7].
It is deeply regrettable for me to observe the decline in academic performance of university
students in these 7 to 8 years. What is meant by the term “academic performance” here is
an ability to think out a problem by oneself rather than the amount of knowledge one has. In
short, it is what we call wisdom. Since the economy of Japan, which lacks natural resources,
can find its way to survive only by relying on the creativity of the Japanese in science and
technology, this decline in academic performance of students must be taken seriously as a
threat to the future of Japan.
Prof. Kodaira also argued as follows. The talent for mathematics, namely, the
talent to think mathematics and understand it, is not quite the same as brightness in
the generic sense, and it varies greatly from one person to another. It is as though
we need to introduce a new kind of human sense to account for such a situation.
We shall call this math-sense. Like the sense of direction, there is no particular
organ of the body that produces math-sense. The inborn level of math-sense varies
considerably and hence not every person is endowed with it or with the same level
of math-sense. This may be because math-sense grew up at quite a later stage of
human evolution. (Prof. Kodaira further told me informally that, in ancient times,
those who were too weak-sighted to distinguish their mother from a wolf must have
died out due to natural selection. On the contrary, those with weak math-sense
were not wiped out because math-sense did not make much difference in ancient
environments.) As the individual math-sense is considerably different among us,
we cannot justify the so-called uniform principle to be applied to mathematics
education.
Moreover, even if one is endowed potentially with good math-sense, it will not
be awakened unless he/she continues proper studying of mathematics earnestly from
childhood. In this connection, we must enrich the language education in elementary
schools as a prerequisite for a good start toward math-sense. This was the claim of
Prof. Kodaira, which I support fully.
I feel that the concept of math-sense is useful as a model that helps us study the
manifestation and development of the ability to think mathematically.
I propose that high school education should aim at awakening the math-sense of
students and letting them acquire the habit of thinking mathematics deeply, without
discriminating between science-oriented students and the others. Moreover, people
in charge of evaluation of students’ scholastic achievements should try to invent an
appropriate way to appreciate (i.e., diagnose the growth of math-sense of) students
who have done such proper study of mathematics and have reinforced the ability to
think mathematically. This kind of effort would be valuable from the point of view
to train future business leaders endowed with mathematical intelligence.
Mathematics in Liberal Arts Education
In Japan before World War II, the entrance into leading national universities
(imperial universities) was practically limited to graduates of high schools under
the prewar educational system. According to this educational system, the total quota
of the imperial universities equaled the total quota of high schools. In other words,
the number of high schools was so limited under the prewar education system. In
those days, liberal arts education used to be given at high schools as one of the two
backbones together with the character building education, and it was the dominant
part of elite education then.
Under the postwar education system in Japan, although liberal arts education is
supposed to be given in the first year of university normally, it currently shows a
tapering off trend due to pressure from hectic professional education. Nevertheless,
as the bulk of professional education is being shifted to graduate schools these
days, some people even go as far as to argue that liberal arts education should be
expanded to cover the entire period of undergraduate studies, while modern liberal
arts education must keep in touch with ideas and methodology at the fronts of
academic fields. (The essay of President Kohichi Hori already cited is one of those
voices.)
In liberal arts education today, mathematics plays an important role. It was
the case in the old days as well; originally, of the seven subjects constituting the
classic liberal arts, more than half of them (arithmetic, geometry, logic, grammar,
rhetoric) were related to mathematics, with mathematical literacy to be considered
as a linguistic ability. The other subjects were music and astronomy, while I like to
consider that the latter can be included in mathematical sciences.
Even putting aside such old history, mathematics is still expected to play a
major role in future liberal arts education. That said, mathematical education
expected today for the university level must not limit itself to a blind transmission
of practical methods and techniques, nor should it be too much concerned with
detailed conventions of professional mathematicians. Rather, it should be aimed
at acquisition of mathematical literacy as stated above in the section Mathematics
for Business and Businesspeople, and should help students understand the method-
ology of mathematical sciences as stated in the section Methodology of Applied
Mathematics, thereby allowing them to attain a unified view on the fundamentals
of mathematics and its applications. This knowledge of mathematics can be said
to be a passport to the frontier of our present civilization. Before World War II,
high school students used to eagerly study Western languages such as English and
German as a “passport” that enabled them to access the frontier of knowledge
144 H. Fujita
(although they were not so concerned with the ability to communicate orally). I think
that, nowadays, acquiring mathematical intelligence plays the role of a “passport”
for all students, whether they major in science or not, that paves the way toward
international leadership at the forefront of various fields including business.
Concluding Remarks
Mathematics for business is an application of mathematics aimed at practical

purposes. The manner in which businesspeople utilize mathematics for business
affairs and management decisions, through cultivated mathematical literacy and
mathematical thinking, seems to resemble the manner of working of clinicians
whose daily practice is carried out with the aid of academic knowledge accumulated
through scientific research in fundamental medicine. Conversely, when a profes-
sional mathematician wishes to contribute to the scientific part of mathematics
for business, he/she should be determined to assume an attitude similar to that
of a medical scientist who is dedicated to academic research and eagerly wishes
to contribute to clinical treatments. Although not all mathematicians need to be
engaged in it, I believe the mathematics community as a whole must hold to a
respectful understanding and encouragement of mathematics for business as long
as the slogan of mathematics for society is agreed to.
Reflecting on the key words mathematics for society, I would like here to say that,
in order for mathematics for business in Japan to reach the level of being appreciated
internationally, it must be developed while taking account of the merits and demerits
of Japanese culture. The strong points are to be activated out and the weak points
are to be reinforced.
As can be seen from a linguistic comparison between Japanese and English, the
Japanese mentality is fundamentally more emotional than logical. Let me elaborate
on that. I note that the following two facts are somehow related: the first is that
Greek geometry took up universally evident propositions as the starting axioms. The
second is that the basis of interpersonal relationships in Western society is placed
on contracts. In contrast to this, the normal relationship in traditional Japanese
society depended on emotional faith and trust, due to the historical experience
of being an isolated and racially homogeneous society, particularly, through the
highly stable Edo era when typical Japanese culture with the two components of
samurai and chonin (general citizens) was formed. Direct but silent sympathy from
heart to heart was appreciated as a more ideal form of human communication than
“persuasion by rhetoric.” It might be the way of traditional Japanese mentality to
think that, if a thing is evident to everyone, there is no need to state it explicitly.
(However, the recent Japanese social atmosphere is deviating from the traditional
one, for good or ill.) As a corollary to what is stated above, it is vital for us,
the Japanese, to study mathematics eagerly and deeply in order to acquire rational
mathematical intelligence, which is the respectable essence of and a firm basis of
Western intelligence.
On the other hand, in the modern information society, where human relation-
ships have diversified, it is more necessary than before for leaders of teams and
organizations to be warmhearted with human sensitivity to others. Moreover, the
world of integrated sciences has entered such a phase that “Teamwork can produce
creativity,” as was described by Koichi Tanaka, who received the Nobel Prize in
Chemistry for his research on mass spectroscopy.
I would like to conclude this chapter with the hope that mathematics for
business in Japan will be built and flourish on the arching combination of Western
intelligence and Japanese sentiment, and will contribute much to the world as well
as to our future.
References
1. Fujita H (1966) On the blowing-up of solutions of the Cauchy problem for ut = Δ u + u1+α .
J Fac Sci Univ Tokyo 13:109–124
2. Fujita H (1985) The present state and a proposed reform of mathematical education at senior
secondary level in Japan. J Sci Educ 9(2):39–52
3. Fujita H (1987) The present state and current problems of mathematics education at the senior
secondary level in Japan. In: Development in school mathematics education around the world,
vol. 1. National Council of Teachers of Mathematics and The University of Chicago School
Mathematics Project, Chicago, pp 191–224
4. Fujita H (2004) Goals of mathematical education and methodology of applied mathematics. In:
Proceedings of the 9th international congress on mathematical education, Tokyo/Makuhari, 31
July–6 August 2000. Kluwer Academic, Dordrecht, pp 19–36
5. Fujita H (2007) A suggestion to statistics education on its spirit and approach—from viewpoint
of mathematics education. Nihon Tōkei Gakkai shi 36(2):219–230 (in Japanese)
6. Giga Y, Kohn RV (1985) Asymptotically self-similar blowing-up of semilinear heat equations.
Comm Pure Appl Math 38:297–319
7. Kodaira K (1984) Gensoku wo wasureta syotou/chutou kyouiku—nan no tame, soshite dare no
tameni isogunoka. Iwanami Kagaku 54(1):4–9 (in Japanese)
8. Kagaku to gijyutsu no aida, International Institute for Advanced Studies, Kyoto (2000)
(Koutouken sensyo no.7) (in Japanese)
9. Poincaré H (1905) La Valeur de la Science. Flammarion, Paris [English translation by Halsted
GB (1929) The value of science in The Foundations of Science. Science Press]
A Brief Biography of Hiroshi Fujita, Mathematician
Hiroshi Fujita is a Professor Emeritus, the University of Tokyo. He received his BA

and PhD at the University of Tokyo.
Prof. Fujita has served in various important roles for the mathematical commu-
nity over the course of his career, including President of the Mathematical Society
of Japan, President of the Japan Society for Industrial and Applied Mathematics,
and President of the Mathematics Education Society in Japan. At the ICME-9, held
146 H. Fujita
in Makuhari, Japan in 2000, he served as the Chairman of the International Program

Committee as well as the President of the National Organizing Committee. Fujita
was also the founding President of the Japan Mathematical Olympiad.
His academic honors and awards include the Fujihara Award, the Medal with
Purple Ribbon and The Order of the Sacred Treasure, Gold and Silver Star.

What Mathematics Can Do For You

Загружено:

Сведения о документе

Авторское право

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

What Mathematics Can Do For You

Загружено:

Авторское право:

Yoshikazu Giga · Toshiyuki Kobayashi

ISBN 978-4-431-54345-9 ISBN 978-4-431-54346-6 (eBook)

© Springer Japan 2013

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Tokyo, Japan Yoshikazu Giga

A Dialogue Between a Mathematician and Toyota’s Chairman:

Iwane Laboratories, Ltd.: Mathematics in Our Company:

Fujio Cho Toshiyuki Kobayashi

What Lies Beneath

If mathematics could be useful in the real world, how? Since I am a mathemati-

(Mathematical) Reason and Sentiment

C: Reason is antithetical to sentiment. It is very important to use rationality while

Mathematics and “Kaizen”

Ask “Why” Five Times

Company Profile of Toyota

Toyota Motor Corporation was founded in 1937. The company’s headquarters is

A Brief Biography of Fujio Cho, Toyota

About the University of Tokyo

Known by his contributions to education and research in Japan is Dairoku

A Brief Biography of Toshiyuki Kobayashi, Mathematician

Toshiyuki Kobayashi, a mathematician, was born in 1962. When he was 25, he

Reasons Why Mathematics Is Important to Our Company

When trying to demonstrate the necessity of mathematics in finance, I usually begin

X(τ ) = (1 + rτ · τ )X(0) (1)

X(nτ ) = (1 + R0→τ · τ )(1 + Rτ →2τ · τ ) · · · (1 + R(n−1)τ →nτ · τ )X(0). (4)

X(0.5y) = (1 + R0→0.5y · 0.5y)X(0), (4)

X(1y) = (1 + R0→0.5y · 0.5y)(1 + R0.5y→1y · 0.5y)X(0), and (4)

X(1.5y) = (1 + R0→0.5y · 0.5y)(1 + R0.5y→1y · 0.5y)(1 + R1y→1.5y · 0.5y)X(0). (4)

X(0) = X(nτ ) · DFnτ . (6)

If both sides are divided by X(0), this equation becomes:

The Market Convention Equation

Secondly, the annual yield pn gives the following equation:

1 = pn DF1y + pn DF2y + · · · + pn DF(n−1)y + (1 + pn)DFny . (16)

Therefore pn is given by:

If we assume (13) holds, by substitution in (15) and (17) we obtain:

Reasons Why Business Needs Mathematics

I would like to refer back to the Eq. (8):

DF1y + DF2y + · · · + DFny < DF0.5y + DF1.5y + · · · + DF(n−0.5)y. (21)

2(DF1y + DF2y + · · · + DFny ) < DF0.5y + DF1y + DF1.5y + · · · + DF(n−0.5)y + DFny.

Next I divide both sides by 2(1 − DFny)(> 0) to yield

DF1y + DF2y + · · · + DFny DF0.5y + DF1y + DF1.5y + · · · + DF(n−0.5)y + DFny

Using (15) and (17), you get

DF0.5y + DF1.5y + DF2.5y + · · · + DF(n−0.5)y < 1 + DF1y + DF2y + · · · + DF(n−1)y.

As before, you can deduce that

Company Profile of BNP Paribas

BNP Paribas is a world-leading banking group headquartered in Paris. As of July

A Brief Biography of Yusuke Yasuda, BNP Paribas/Tokyo

Are Numbers Useful?

The Realities of Corporate Management

contributions are made by quantitative factors. Here, mathematical modeling, which

The Mathematical Methods Employed by Takeda

Takeda is a pharmaceutical company. Like most companies, it has a number of

Corporate Strategy and Planning

Sales and Marketing

Research and Development

Research and development is of the utmost importance for pharmaceutical compa-

hypothesis or using mathematical reasoning as the basis for decision making. It is

Mathematical Reasoning in Business

Last But Not Least

Company Profile of Takeda Pharmaceutical Company Limited

Founded in 1781, Takeda Pharmaceutical Company Limited is a leading Japanese

A Brief Biography of Yasuchika Hasegawa, Takeda Pharmaceutical

Strategy & Planning Department. He also serves as Chairman of the Keizai

Mathematics Drives the Economy

thinking mentioned above is very helpful in solving problems in such situations. In

Mathematics Supports the Safety and Security of the Services