Вы находитесь на странице: 1из 4

Digitalisation as a strategic choice at University

Challenges in research and education


Ivica Crnkovic and Gordana Dodig-Crnkovic
Chalmers University of Technology, Gothenburg, Sweden
ivica.crnkovic@chalmers.se, gordana-dodig.crnkovic@chalmers.se
Abstract. The basis of digitalisation comes from ICT, but it applies
to variety of different domains, with their domain specific knowledge,
practices and methods. In this paper we present some of challenges of
digitalisation of research and education in the domain of dependable
cyber-physical systems, from the perspective of Chalmers University of
Technology.
Keywords: digitalisation, cyber-physical systems, dependability

Digitalisation - a new approach

Digitalisation: Integration of digital technologies into everyday life by the digitisation of everything that can be digitised. (businessdictionary.com 2014). It
includes a process of transforming artefacts, relations, processes, methods, and
theories into a digital form suitable for use by computers, where advantages of
information processing technologies can be utilised. Digitalisation has potential
to revolutionise the development, manufacturing, and use of the products of today; it can support radically new product features, enable new products and
services, and generate new business opportunities that were not even imaginable
before[1]. Many countries have prioritised digitalisation as a key enabler for the
industrial and societal development1 . For example, both Sweden and Denmark
have a Ministry of Digitalisation. While the digitalisation process is already in
full exploitation in many areas, there are significant challenges related to how to
develop and utilise all its advantages and how to manage the emerging risks. Due
to an overall enthusiasm and a tremendous speed in digitalisation, connected to
non-critical acceptance, there is a possibility that, before it comes to its stable
phase, digitalisation process can pass through a turbulent period, with risks for
financial losses, lost jobs with social consequences and serious security and safety
threats[2]. For this reason it is necessary that different societal structures have
knowledge and responsibility for safe and successful digitalisation process, which
will benefit both individuals, and the whole society.
The fundamentals of digitalisation lie in Informatics, in which processes,
organisation, and management of the domain-specific knowledge are abstracted
and transformed into models, data structures and algorithms. The main point
in digitalisation is utilisation of the ability of Informatics to handle the domain
processes, and to integrate digital features with the domain artefacts. On one
hand this requires changes in the domain artefacts and processes, and on the
1

EC for Digital Jobs (2014), http://issuu.com/digitaleurope/docs/grand coalition10/c/sl8989x

other hand new requirements are put on Informatics. In these complex relations
it is challenging to find out which forms and methods from Informatics are
fundamental for digitalisation and becoming standard digitalisation components
in different application domains.
In this position paper we give an overview of digitalisation concerns in the
research and development of dependable cyber-physical systems (CPS) for automotive industry, in relation to university research and education at Chalmers.

Digitalisation for dependable CPS - research directions

Development of complex dependable CPS is usually costly and a long time-tomarket process, as the dependability assurance (of safety, reliability, security,
etc.) requires significantly more resources than functional development alone.
Through digitalisation this process can be improved with respect to the leadtime, increased functionality and product variability, and increased dependability characteristics, including resilience due to dynamic adaptation[3]. Research
in digitalization at Chalmers related to dependable CPS is done in cooperation
with several vehicular companies, including Volvo and Autoliv. The following
research directions have been identified.
Model- and software-driven approach. Since most of design artefacts can be
digitised in form of different models, majority of the development can be done
on a model-level, with possible automatic generation of the production artefacts. Digitalisation enables creation of many model variants and their tests in
a simulation environment. Further, a lot of information can be collected from
the products in their use, including the user experience providing a continuous feedback that enables a continuous experimentation, product improvements
and deployment. A continuous development and deployment removes a border
between the development, production, and utilisation. The challenges imposed
here are related to the system consistency, since many component variants can
violate the system integrity.
Data-driven approach. The new ICT technologies and e-infrastructure have
enabled accessing and processing larger amounts and diversity of data. A system can use data generated in the system, data from the environment, but also
from other sources. In addition, new technologies such as deep learning enable
generation of new knowledge from the existing data. The provided information
can be used in the development process for a better and more effective validation, or experimentation to achieve a better system performance and increased
customer value. Further, data collected with the services can be used in real
time to dynamically adapt the system to the new requirements or a new context. Deep learning and other machine learning techniques provide new functions
and significantly better accuracy, flexibility, and adaptability. Extensive use of
data in combination with machine learning is a new paradigm in system development. This also imposes new challenges in the development process and in
ensuring the system dependability. A big challenge is how to effectively process
large amounts of data, in particular in real time, as data storage, data communication, and data analysis are determined by limited resources. Hierarchical

and non-relational databases are being developed to increase the efficiency in


data management, and new heterogeneous execution platforms with different
computational characteristics such as high performance and high parallelism,
are becoming standard parts of dependable CPS. Another type of challenge is
related to heterogeneity of data, including its accuracy, reliability, and trustworthiness. As a consequence of the fact that these systems are becoming highly
connected, new challenges related to security and safety appear.
Value-driven approach. Basic values are driving the development with the
aim to improve human condition. However, digitalisation can have severe consequences for employment in both the short or in the long run, which in its turn
may lead to significant problems for human well-being, social structures and
for the democracy. The question of sustainability of new approaches and strategies that come with digitalisation is of prime importance. At the same time,
the digitalization can enable new business opportunities and new applications.
For example, with autonomous cars many new services are expected, such as
transport on demand of both people and goods, or new forms of traffic control.
New business opportunities will generate cooperation of different domains that
previously had no mutual relations. An example is a cooperation of Tesla, the
electrical cars manufacturer, and SolarWorld AG, the solar cells manufacturer.
This also leads to new challenges, such as issues related to IPRs and accessibility
of information between different companies.

Interdisciplinarity of research for digitalisation

Digitalisation research cuts across traditional research domains and has high
demands on interdisciplinary research. Chalmers strategically supports interdisciplinary research through Area of Advances (AoA2 , organisational entities
complementary to university departments, and strategically chosen as challengedriven thematic platforms relevant for industry and society. AoA ICT3 drives,
through the ICT research groups the fundamental research in Informatics, including the digitalisation-related topics, and provides a support in digitalisation
to other AoAs. Addressing the paradigm of digital world AoA ICT has dened the
following four challenge-oriented profiles in which the digitalisation challenges
addressed by the research.
Automated Society - ICT enabling autonomous systems and intelligent automation of processes. Here specifically, a research topic Digitalisation in dependable CPS development is addressing integration of software-inspired processes
and methods in the system development including agile system development,
continuous integration and deployment, and dynamic software product lines.
Connected World - Advanced communication and sensors, including new
technologies in algorithms, hardware, software, and systems. The research challenge of CPS addressed here is Efficient data processing and communication that
deals with intelligent ways to process large amounts of data being transferred
between different units in a distributed computing system.
2
3

AoA http://www.chalmers.se/en/areas-of-advance/Pages/default.aspx
AoA ICT http://www.chalmers.se/en/areas-of-advance/ict/about us/Pages/default.aspx

Big Data -Managing large amounts of data, advanced computation and modelling based on data. The CPS research topic addressed is Evidence-based System Adaptation. It targets system on-line adaption and machine-learning analysis
based on run-time data, accumulated data, and experience, using the latest technologies of heterogeneous execution platforms, and targeting specific problems,
such as 3D vision and recognition
Digital Sustainability - Sustainable development including environmental,
economic, social and technical aspects with the aid of ICT technologies. For the
CPS, here are addressed Digital Safety and Security in the context of intensive
data exchange and dynamic system adaption.

Educational framework for digitalisation

Digitalisation brings new approaches and new agendas not only to research,
but by necessity also affects the other central university function - education.
The Chalmers University of Technology has started an initiative in cooperation
with the University of Gothenburg and several industrial partners in order to
prioritise digitalisation in the education. Data Science has been selected as a
strategically important field. On the masters level, an initiative is to identify
and adapt courses from relevant masters programs (from Mathematics, ICT
programs, and Biotechnology) for the master students from different programs,
who obtain given number of credits to get a certificate of Data Science engineer.
In parallel, through the cooperation, a new master program in Data Science
is initiated. This program has a broader choice of courses - from fundamental
courses related to databases, data analytics and machine learning, to selective
courses, projects, and work on master thesis from other programs.
Graduate education at Chalmers is organised at the departments, but in
special cases Chalmers organises or takes part in research schools. A research
school gathers a group of PhD students and organises special events for them,
such as common courses, summer schools, visits to research and industrial centres, workshops, and possibly research projects. There is on-going discussion to
build up such a research school in digitalisation. Led and funded by ICT AoA
this school is supposed to gather 10-20 PhD students from different fields (ICT
fields, mathematics, biotechnology, architecture, production, etc.). The challenge
for this school will be to identify common courses, and possibly common research
project through which the PhD students could build shared interdisciplinary
foundations for digitalisation and its application to their research fields.
We can conclude that there are big promises and challenges ahead, both in
research and education, but research seems to be a step ahead of the education.

References
1. Brynjolfsson, E., McAfee, A. The second Machine Age: Work, Progress and Prosperity in a time
of brilliant Technologies. W. W. Norton & Company Inc. New York, 2014.
2. Yoo, Y. et al.: Organizing for Innovation in the Digitized World. Organization Science, 23, 2012.
3. E. A. Lee, Cyber Physical Systems: Design Challenges, 2008 11th IEEE International Symposium on Object and Component-Oriented Real-Time Distributed Computing (ISORC), Orlando,
FL, pp. 363-369, 2008.

Вам также может понравиться