Corrupting Influence

of Variability
When luck is on your side, you can do
without brains.
Giordano Bruno, burned
at the stake in 1600

The more you know the luckier you get.

J.R. Ewing of Dallas
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

Performance of a Serial Line

Measures:

Links to Business Strategy:

Throughput
Inventory (RMI, WIP, FGI)
Cycle Time
Utilization
Lead Time
Customer Service
Quality

Evaluation:

Would inventory reduction result in

significant cost savings?
Would CT (or LT) reduction result
in significant competitive
advantage?
Would TH increase help generate
significantly more revenue?
Would improved customer service
generate business over the long
run?

Comparison to perfect values

(e.g., rb*, T0*)
Relative weights consistent
with business strategy?

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

Influence of Variability
Variability Law: Increasing variability always degrades the

performance of a production system.

Examples:

process time variability pushes best case toward worst case

higher demand variability requires more safety stock for same level of
customer service
higher cycle time variability requires longer lead time quotes to attain
same level of on-time delivery

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

Variability Buffering
Buffering Law: Systems with variability must be buffered by

some combination of:

1. Inventory,
2. Capacity, or
3. Time

Interpretation: If you cannot pay to reduce variability, you will pay in

terms of high WIP, under-utilized capacity, or reduced customer service
(i.e., lost sales, long lead times, and/or late deliveries)

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

Variability Buffering Examples

Ballpoint Pens:
cant buffer with time (who will backorder a cheap pen?)
cant buffer with capacity (too expensive, and slow)
must buffer with inventory

Ambulance Service:
cant buffer with inventory (stock of emergency services?)
cant buffer with time (violates strategic objectives)
must buffer with capacity

Organ Transplants:
cant buffer with inventory (perishable)
cant buffer with capacity
must buffer with time

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

Push vs. Pull: Corruptive Influence

Which type of production control is more conducive to
corruptive-influence-of-variability: Push or Pull?

What are the performance tradeoffs between these two

production control alternatives?
PUSH

ra

ca

1
B (1)=

TH Constrained

PULL

Infinite
raw
materials

WIP Constrained

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

te(1)
ce(1)

2
B (2)=

1
te(1) B (2)<<
ce(1)

te(2)
ce(2)

3
B (3)=

2
te(2)
ce(2)

te(3)
ce(3)

4
B (4)=

3
B (3)<<

te(3)
ce(3)

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

te(4)
ce(4)

4
B (4)<<

te(4)
ce(4)

Corruptive Influence of Var. in Push Systems

Case
1
2
3
4

te (i )
i = 1, 2, 4 te (3)
(min)

(min)

1
1
1
1

1.2
1.2
1.0
1.2

c (i )
i =1-4
0
1
1
0.3

TH

(j/min)

CT (min)

WIP

(jobs)

CT

(min)

Comments

System Parameters:
ra = 0.8, ca = ce (i ) in all
cases
B (i ) = , i = 1-4 in all cases

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

Corruptive Influence of Var. in Pull Systems

te (i )
CT
i = 1,2,4 te (3) c (i )
B (3)
TH
CT
WIP
Case
(min)
(min) i = 1-4 (jobs) (j/min) (min) (jobs) (min) Comments
1
1
1.2
0
0
2
1
1.2
1
0
3
1
1.2
1
1
4
1
1.2
0.3
0
5
1
1.2
0.3
1
6
1
1.2
0.3
0
System Parameters:
Station 1 pulls in job whenever it becomes empty
B (i ) = 0, i = 1, 2, 4 in all cases, except case 6, which
has B (2) = 1

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

Corruptive Influence of Var. in Push Systems

Simulation Results
f(ra , rb)
te (i )

c (i )

i = 1, 2, 4

te (3 )

( m in )

( m in )

i = 1 -4

( j/ m in )

1 .2

0 .8

1 .2

1 .0

1 .2

C ase

W IP

CT

( jo b s )

( m in )

4 .2

3 .4

0 .0

0 .8

4 4 .6

3 5 .7

2 6 .8

W I P B u ffe r

0 .8

2 0 .0

1 6 .0

1 0 .3

C a p a c ity B u ffe r

0 .3

0 .8

7 .8

6 .2

3 .3

TH
CT

( m in )

C o m m e n ts

B est C ase

R e d u c e d V a ria b ility

System Parameters:
ra = 0.8, ca = ce (i ) in all cases
B (i ) = , i = 1-4 in all cases

Observations:
TH is set by release rate in a push system
Increasing capacity (rb) reduces need for WIP buffering
Reducing process variability reduces WIP, CT, and CT variability
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

Corruptive Influence of Var. in Pull Systems

Simulation Results
te (i )
i = 1,2,4
Case (min)
1
1
2
1
3
1
4
1
5
1
6
1

te (3) c (i ) B (3)
(min) i = 1-4 (jobs)
1.2
0
0
1.2
1
0
1.2
1
1
1.2 0.3
0
1.2 0.3
1
1.2 0.3
0

TH
(j/min)
0.83
0.48
0.53
0.72
0.76
0.73

CT
(min)
4.2
6.4
7.2
5.0
6.0
6.3

WIP
(jobs)
3.5
3.1
3.8
3.6
4.5
4.6

CT

(min)
Comments
0.0
Best Case
2.4
Plain JIT
2.6
Inv Buffer
0.6
Var Reduction
0.8 Inv Buffer + Var Reduction
0.7
Non-bottleneck Buffer

System Parameters:
Station 1 pulls in job whenever it becomes empty
B (i ) = 0, i = 1, 2, 4 in all cases, except case 6, which
has B (2) = 1
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

10

Corruptive Influence of Var.: Observations

Buffering law applies to both cases

Consequences different in push and pull systems

Capping WIP without reducing variability reduces TH

WIP cap limits effect of process variability on WIP/CT
Reducing process variability increases TH (given same WIP)
Adding buffer space at bottleneck most increases TH
Impact of buffer size depends on variability
Reducing process variability reduces CT variability

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

11

Buffer Flexibility
Buffer Flexibility Corollary: Flexibility reduces the amount

of variability buffering required in a production system

Examples:
Flexible Capacity: cross-trained workers
Flexible Inventory: generic stock (e.g., assemble to order)
Flexible Time: variable lead time quotes

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

12

Variability from Batching

VUT Equation:
CT depends on process
variability and flow
variability

Batching:
affects flow variability
affects waiting inventory

Conclusion:
Batching is an important
determinant of performance
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

13

Process Batch vs. Move Batch

Process Batch:
Related to length of setup
The longer the setup the larger the lot size required for the
same throughput

Move (transfer) Batch:

The smaller the move batch, the shorter the cycle time
The smaller the move batch, the more the material handling

Lot Splitting:

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

Move batch can be different from process batch

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

14

Process Batching Effects

Types of Process Batching:
Serial Batching:
processes with sequence-dependent setups
batch size is number of jobs between setups
batching used to reduce loss of capacity from setups

Parallel Batching:
true batch operations (e.g., heat treat)
batch size is number of jobs run together
batching used to increase effective rate of process

Process Batching Law:

In stations with batch operations:

The minimum process batch size that yields a stable system may be
greater than one
As process batch size becomes large, CT grows proportionally with
batch size
Batch CT at the station will be minimized for some process batch size,
which may be greater than one

Basic Batching Tradeoff: WIP (or CT) vs. throughput (utilization)

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

15

Serial Process Batching Example

Parameters:
k

k s e r ia l b a t c h s iz e ( 1 0 )
t t im e t o p r o c e s s a s in g le p a r t ( 1 )
s t im e t o p e r fo r m a s e t u p ( 5 )
c e C V fo r b a t c h ( p a r t s s e t u p ) ( 0 .5 )
Assumption:
Independent of k

ts
setup

t0

ra
ca

r a a r r iv a l r a t e fo r p a r t s ( 0 .4 )

forming
batch

c a C V o f b a t c h a r r iv a ls ( 1 .0 )

queue of
batches

Time to process batch:

t e ( b a tc h ) k t s
1 0 (1 ) 5 1 5

Setup is part of
batch cycle time!

Arrival of batches:
r a ( b a tc h ) r a / k
0 .4 / 1 0 0 .0 4

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

16

Serial Process Batching Example (cont.)

Utilization:
u r a ( b a tc h ) t e ( b a tc h )

or

r a (t s / k )

0 .0 4 (1 5 ) 0 .6

u is defined as

effective utilization
(increases with s)

For stability: u < 1 requires

k

sra

minimum batch
size required
for stability of
system...

1 tra
5(4)
1 1 ( 0 .4 )

Not Contradictory:
Explain Why?

3 .3 3

sra
1 tr a

sra
1 tr a

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

L o w u b u t h ig h C T ( h ig h T H )

L o w C T b u t h ig h u ( lo w T H )

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

17

Serial Process Batching Example (cont.)

Average queue time at station:
2

ca ce u

C Tq

te
2
1

1 0 .5 0 .6

1 0 .6

1 5 1 6 .8 7 5

Note: we assume
arrival CV of
batches is ca
regardless of
batch size an
approximation...

Average cycle time depends on move batch size:

Move batch = process batch

C T n o n s p lit C T q t e C T q s k t
1 6 .8 7 5 1 5 3 1 .8 7 5

Note: splitting move

batches reduces cycle time

Move batch = 1
C T s p lit C T q s

k 1
2

1 6 .8 7 5 1 0

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

10 1
2

(1 .0 ) 2 7 .3 7 5

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

18

(Batch) Cycle Time vs. Batch Size 5 hr setup

80.00
70.00

Cycle Time (hrs)

60.00
50.00
40.00
No Lot Splitting

30.00

Lot Splitting

20.00
10.00

0.00
0

10

Optimum
Batch Sizes
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

15

20

25

30

35

40

45

50

Batch Size (jobs/batch)

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

19

Cycle Time vs. Batch Size 5 hr setup NEW

Eqns.
80
No Lot Splitting

Cycle Time (hrs)

70

Lot Splitting

60

No Lot Splitting - NEW

Lot Splitting - NEW

50

40
30
20
10
0
0

10

15

20

25

30

35

40

45

50

Batch Size (jobs/batch)

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

20

Cycle Time vs. Batch Size 2.5 hr setup

80.00

Cycle Time (hrs)

70.00

60.00
50.00
40.00
No Lot Splitting

30.00

Lot Splitting

20.00
10.00
0.00
0

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

5
10
Optimum
Batch Sizes

15

20

25

30

35

40

45

50

Batch Size (jobs/batch)

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

21

Setup Time Reduction

Where?
Stations where capacity is expensive
Excess capacity may sometimes be cheaper

Steps:
1. Externalize portions of setup
2. Reduce adjustment time (guides, clamps, etc.)
3. Technological advancements (hoists, quick-release, etc.)

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

22

Parallel Process Batching Example

Parameters:
k p a r a lle l b a t c h s iz e ( 1 0 )

t t im e t o p r o c e s s a b a t c h ( 9 0 )
c e C V fo r b a t c h ( 1 .0 )
r a a r r iv a l r a t e fo r p a r t s ( 0 .0 5 )
c a C V o f p a r t a r r iv a ls ( 1 .0 )

ra
ca

forming
batch

B m a x im u m b a t c h s iz e ( 1 0 0 )

queue of
batches

Time to process batch:

Arrival of batches:

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

23

Parallel Process Batching Example (cont.)

Utilization:
u r a ( b a tc h ) t e ( b a tc h )

o r ( r a / k )t

0 .0 0 5 ( 9 0 ) 0 .4 5

For stability: u < 1 requires

k rat
0 .0 5 ( 9 0 )
4 .5

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

minimum batch
size required
for stability of
system...

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

24

Parallel Process Batching Example (cont.)

Average wait-for-batch time:

WT

k 1 1
2

ra

10 1

0 .0 5

90

batch size affects

both wait-for-batch
time and queue time

Average queue plus process time at station:

2

ca /k ce u

0 .1 1 0 .4 5
CT
t

9 0 9 0 1 3 0 .5
2
1

u
2
1

0
.4
5

Total cycle time:

C T W T 9 0 1 3 0 .5 2 2 0 .5

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

25

(Batch) Cycle Time vs. Batch Size in a Parallel

Operation

1400.00

Total Cycle Time

1200.00

wait for
batch time

queue time due

to poor
utilization

1000.00
800.00
600.00
400.00
200.00
0.00
0
10
20
Optimum
Batch Size

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

30

40

50

Nb

60

70

80

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

90

100 110
B

26

Move Batching
Move Batching Law: Cycle times over a segment of a

routing are roughly proportional to the transfer batch

sizes used over that segment, provided there is no
waiting for the conveyance device.

Insights:
Basic Batching Tradeoff: WIP vs. move frequency
Queuing for conveyance device can offset CT reduction from
reduced move batch size
Move batching intimately related to material handling and layout
decisions

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

27

Move Batching Example

Problem:
Two machines in series
First machine receives individual parts at rate ra with CV of ca(1)
and puts out batches of size k
First machine has mean process time of te(1) for one part with
CV of ce(1)
Second machine receives batches of k and puts out individual
parts
Second machine has mean process time of te(2) for one part
with CV of ce(2)
How does cycle time depend on the batch size k ?
ra,ca(1)

te(1),ce(1)

te(2),ce(2)

single job
batch

Station 1
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

Station 2

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

28

Move Batching Calculations

Move Batching Calculations

Time at First Station:

Average time before batching is:

2

c a (1 ) c e (1 ) u (1 )

1 u (1 )

Average time forming the batch is:

k 1 1
ra

regular VUT
equation...

t e (1 ) t e (1 )

k 1
2 u (1 )

first part waits (k-1)(1/ra),

last part doesnt wait, so
average is (k-1)(1/ra)/2;
ra(1)= rd(1) to achieve
conservation of flow

t e (1 )

Average time spent at the first station is:

2

C T (1 )

c a (1 ) c e (1 )

u (1 )

1 u (1 )

C T ( 1 , n o b a t c h in g )

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

t e (1 ) t e (1 )
k 1
2 u (1 )

k 1
2 u (1 )

t e (1 )

t e (1 )

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

29

Move Batching Calculations (cont.)

Output of First Station:

Time between output of individual parts into the batch is ta.

Time between output of batches of size k is kta.

Variance of interoutput times of parts is cd2(1)ta2, where

2

because
cd2(1)=d2/ta2
by def of CV

c d (1 ) (1 u (1 ) ) c a (1 ) u (1 ) c e (1 )

Variance of batches of size k is kcd2(1)ta2.

SCV of batch arrivals to station 2 is:

2

c a (2 )

k c d (1 )t a
2

k ta

because departures are

independent, so variances add
variance divided by
mean squared...

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

c d (1 )
k

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

30

Move Batching Calculations (cont.)

Time at Second Station:

independent
process times...

Time to process a batch of size k is kte(2).

Variance of time to process a batch of size k is kce2(2)te2(2).
Squared CV for a batch of size k is:
2

k c e ( 2 )t e ( 2 )
2

k t e (2 )

Mean time spent in partial batch of size k is (broken batch):

k 1
2

c e (2 )

first part doesnt wait, last

part waits (k-1)te(2), so
average is (k-1)te(2)/2

t e (2 )

So, average time spent at the second station is:

2

c d (1 ) / k c e ( 2 ) / k u ( 2 )
k 1
C T (2 )
t e (2 ) t e (2 )

k t e (2 )
2
1

u
(
2
)
2

VUT equation to
compute queue time
of batches...

c d (1 ) c e ( 2 ) u ( 2 )

1 u (2 )
C T ( 2 , n o b a t c h in g )

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

k 1
t e (2 )
t e (2 ) t e (2 )
2

k 1
2

t e (2 )

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

31

Move Batching Calculations (cont.)

Total Cycle Time:

C T (b a t c h in g ) C T ( n o b a t c h in g )

k 1
2 u (1 )

t e (1 )

k 1
2

t e (2 )

k 1 t e (1 )
C T ( n o b a t c h in g )

t
(
2
)
e

2 u (1 )

inflation factor due

to move batching

Insight:

Cycle time increases with k

Inflation term does not involve CVs
Congestion from batching is more of bad control than

randomness

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

32

Assembly Operations
Assembly Operations Law: The performance of an

assembly station is degraded by increasing any of the

following:
1.
2.
3.

Number of components being assembled

Variability of component arrivals
Lack of coordination between component arrivals

Observations:
This law can be viewed as special instance of variability law.
Number of components affected by product/process design.
Arrival variability affected by process variability and production
control.
Coordination affected by scheduling and shop floor control.

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

33

Cycle Time
Definition (Station Cycle Time): The average cycle

time at a station is made up of the following

components:

cycle time = move time + queue time + setup time

+ process time + wait-to-batch time
+ wait-in-batch time + wait-to-match time

delay times
typically
make up
90% of CT

Definition (Line Cycle Time): The average cycle

time in a line is equal to the sum of the cycle times at

the individual stations less any time that overlaps two
or more stations

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

34

Attacking Variability
Objectives
reduce cycle time
increase throughput
improve customer service

Levers

reduce variability directly

buffer using inventory
buffer using capacity
buffer using time
increase buffer flexibility
Flexible Capacity: cross-trained workers
Flexible Inventory: generic stock
Flexible Time: variable lead time quotes

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

35

Reducing Queue Delay

CTq = V U t

ca ce

Reduce Variability
failures
setups
uneven arrivals, etc.

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

Reduce Utilization

arrival rate (yield, rework, etc.)

process rate (speed, time,
availability, etc)

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

36

Reducing Batching Delay

CTbatch = delay at stations + delay between stations

Reduce Process Batching
Optimize batch sizes

Reduce Move Batching

Move more frequently

Reduce setups
Stations where capacity is
expensive
TH vs. WIP/CT tradeoff

Layout to support
material handling (e.g.,
cells)

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

37

Reducing Matching Delay

CTbatch = delay due to lack of synchronization

Reduce Variability
High utilization
fabrication lines
Usual variability
reduction methods

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

Improve Coordination
scheduling
pull mechanisms
modular designs

Reduce Number
of Components
product redesign
kitting

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

38

Increasing Throughput

TH = P(bottleneck is busy) bottleneck rate

Reduce Blocking/Starving
buffer with inventory (near

bottleneck)
reduce system desire to queue

Increase Capacity
add equipment
increase operating time (e.g.
spell breaks)
increase reliability
reduce yield loss/rework

CTq = V U t
Reduce Variability

Reduce Utilization

Note: if WIP is limited, then system degrades

via TH loss rather than WIP/CT inflation

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

39

Customer Service
Elements of Customer Service:
lead time
fill rate (% of orders delivered on-time)
quality

Law (Lead Time): The manufacturing lead time for a routing that

yields a given service level is an increasing function of both the mean

and standard deviation of the cycle time of the routing.

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

40

Improving Customer Service

LT = CT + z
CT
Reduce CT Visible
to Customer
postponement
(delayed
differentiation)
assemble-to-order

Reduce Average
CT
queue time
batch time
match time

Reduce CT Variability
generally same as
methods for reducing
average CT:
improve reliability
improve maintainability
reduce labor variability
improve quality
improve scheduling, etc.

Assumptions:
- No order back-log (new order can be scheduled without delay)
- Availability of all necessary raw materials

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

41

Cycle Time and Lead Time

0.18
0.16

Lead Time = 14 days

0.14
Densities

0.12
0.1

Lead Time = 27 days

0.08
0.06
0.04
0.02
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40

Cycle Time in Days

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

42

Procoat Case: Diagnostics Using Factory Physics

IN-CLASS ACTIVITY

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

43

Procoat Case: Background

Situation:
Line not meeting goal of 2.4 jobs/hour
Expose has highest utilization: bottleneck station

Proposal:
Install second line at expose station: Expensive
Very little space (clean room needs expansion; expensive task)
Overall, not an acceptable proposal!

Can factory physics help?

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

44

Procoat Case Layout

Scope of Analysis
IN

Loader

Clean

Coat 1

Coat 2

Loader
Stn. #2

Stn. #1

Touchup
Unloader

D&I
Inspect Develop

Bake

E1
Unloader

Manufacturing Inspect
M1

M2

M3

M4

M8

M7

M6

M5

Expose
E5

E2
E3
E4

Clean Room

OUT
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

45

Procoat Case: Details on Key Stations

Two stations in series; Station 2 is bottleneck
Job Arrivals: Ca2 = 1
Station 1: Resist Apply
t 0 1 9 m in
2

c 0 0 .2 5
M T T F 4 8 h r, M T T R 8 h r

Station 2: Expose
t 0 2 2 m in
c

2
0

M T T F 3 .3 3 3 h r , M T T R 1 0 m in

Space for 20 jobs of WIP due to size of clean room

Desired throughput 2.4 jobs/hr, not being met

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

46

Procoat Case: Diagnostics Using Factory Physics

Solution Slides!

Analysis Tools:

Example: Unreliable Machines

f (t 0 )

Intrinsic
Variability

t0

mf mr

te t0 / A

Other Sources
of Variability

mf

f (t e )

ce

te

te

ce (i)
ca (i)

cd (i)
Station

i
te(i), u(i)=ra /re

C Tq
2

2
0

(1 c r ) A (1 A )

ca ce

1 u

t0

VUT equation

te

c d u c e (1 u ) c a

Propagation
Equation

Ask why five times ...

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

47

Procoat Case: Diagnostics Using Factory Physics

Solution Slides!

Analysis:
Step 1: At 2.4 job/hr
CTq at first station is 645 minutes, average WIP is 25.8 jobs
CTq at second station is 887 minutes, average WIP is 35.5 jobs
Space requirements at machine 2 are violated!

Step 2: Why is CTq at machine 2 so big?

Break CTq into
2

C Tq

ca ce u

t e ( 3 .1 6 ) (1 2 .2 2 ) ( 2 3 .1 1 m in )
2

1 u

The 23.11 min term is small (t0/A)

The 12.22 correction term is moderate (u 0.9244)
The 3.16 correction is large

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

48

Procoat Case: Diagnostics Using Factory Phy ...

Solution Slides!

Step 3: Why is the correction term so large?

Look at components of correction term
ce2 = 1.04, ca2 = 5.27
Arrivals to machine are highly variable
Step 4: Why is ca2 to machine 2 so large?
Recall that ca2 to machine 2 equals cd2 from machine 1, and
2

c d u c e (1 u ) c a ( 0 .8 8 7 ) ( 6 .4 3 7 ) (1 0 .8 8 7 ) (1 .0 ) 5 .2 7

ce2 at machine 1 is large

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

49

Procoat Case: Diagnostics Using Factory Phy ...

Solution Slides!

Step 5: Why is ce2 at machine 1 large?

Effective CV at machine 1 is affected by failures,
2

c e c 0 2 A (1 A )

mr
t0

0 .2 5 6 .1 8 6 .4 3

The inflation due to failures is large

Reducing MTTR at machine 1 would substantially improve
performance

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

50

Procoat Case: Diagnostics Using Factory Phy ...

Solution Slides! FORWARD CALCULATIONS

Step A: Estimate te, ce, and cd for machine 1

Machine availability given breakdown pattern:
A

mf

mf mr

48

48 8

0 .8 5 7

Effective cycle times given breakdown pattern:

t e t 0 / A 1 9 0 .8 5 7 2 2 .1 6 7 m in

Effective CV at machine 1 given failure pattern:

2

c e c 0 2 A (1 A )

mr
t0

0 .2 5 2 0 .8 6 7 (1 0 .8 6 7 )

8 60
19

0 .2 5 6 .1 8 6 .4 3

Utilization of machine 1:
u r a / r e r a t e 2 .4 2 2 .1 6 7 6 0 0 .8 8 7

Departure variability at machine 1

2

c d u c e (1 u ) c a (0 .8 8 7 ) (6 .4 3 7 ) (1 0 .8 8 7 ) (1 .0 ) 5 .2 7
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

51

Procoat Case: Diagnostics Using Factory Phy ...

Solution Slides! FORWARD CALCULATIONS

Step B: Estimate CT in the Queue for machine 2

Machine availability given breakdown pattern:
A

mf

mf mr

3 .3 3 3

3 .3 3 3 1 0 / 6 0

0 .9 5 2

Effective cycle times given breakdown pattern:

t e t 0 / A 2 2 0 .9 5 2 2 3 .1 0 0 m in

Effective CV at machine 2 given failure pattern:

2

c e c 0 2 A (1 A )

mr
t0

1 2 0 .9 5 2 (1 0 .9 5 2 )

10 / 60
22 / 60

1 0 .0 4 2 1 .0 4 2

Utilization of machine 2:
u r a / r e r a t e 2 .4 2 3 .1 0 0 6 0 0 .9 2 4

CT in the Queue at machine 2:

C Tq
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

c a2 c e2 u
1
5 .2 7 1 .0 4 2

2 3 .1 1 m in
2
2

1 0 .9 2 4

1 u
IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

52

Procoat Case: Factory Physics Analysis

Expose has highest utilization
increase capacity via break spelling

Expose has high V coefficient (and limited room

for WIP)
some is due to operator variability implement training
program
more is due to arrival variability look at coater line

Coater line has long, infrequent failures

maintain field ready replacements
improve maintenance CTs

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

53

Line Physics: Takeaways

Variance Degrades Performance
Many sources of variability
Planned and unplanned

Variability Must be Buffered

Inventory or Capacity or Lead Time

Flexibility Reduces Need for Buffering

Still need buffers, but smaller ones

Variability and Utilization Interact

Congestion effects multiply
Utilization effects are highly nonlinear
Importance of bottleneck management
WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

54

Line Physics: Takeaways

Batching is an Important Source of Variability
Process and move batching
Serial and parallel batching
Wait-to-batch time in addition to variability effects

Assembly Operations Magnify Impact of

Variability
Function of # parts, variability, coordination

Variability Propagates
Flow variability is as disruptive as process variability
Non-bottlenecks can also be problems

WAYNE
STATE U
R.B.
Chinnam
@NIVERSITY
WSU

IFORS
2011: July
1015,(Source:
2011 - Melbourne,
Australia 2008)
IE7315:
Production
Systems
Hopp & Spearman

55