eEngineering 2009—2012. Digitising the product process

VISIONS
•SCIENCE•TECH
N
OLOGY•RESE
ARCHHIGHLI
GHTS•
eEngineering 2009—2012
In addition to industrial production, the success of Finnish industry is based
strongly on the design and engineering of devices, working machines, manufac-
turing plants, power plants, process machinery and ships for global markets. At
the same time, digitisation has become ever more vital to the success of industrial
production and engineering and the volume, value and importance of the digital,
virtual realm is increasing dramatically compared to physical plants and machines.
At the beginning of the 2010s traditional heavy industry accounted for 75%
of the total value of Finnish exports, up notably from 57% in 2000. To reduce
the design and production ramp-up times by half, VTT’s eEngineering spear-
head programme (2009-2012) developed technology platforms for modelling
and simulation, design knowledge management, life-cycle management, and
human-technology interaction. The highlights of the research carried out during
the programme are presented in this publication.
The most signiﬁcant achievement of the programme is Simantics, an exten-
sive operating system providing an open, high-level application platform on
which different computational tools can be easily integrated to form a common
environment for modelling and simulation. Programme also enabled successful
integration of user’s sound and noise experience and thermal comfort modelling
to design of machine cabins in a virtual the design environment.
ISBN 978-951-38-8125-2 (print)
ISBN 978-951-38-8126-9 (online)
ISSN-L 2242-1173
ISSN 2242-1173 (print)
ISSN 2242-1181 (online)
8
VTTRESEARCHHIGHLIGHTS8eEngineering2009—2012
Digitising the product process

VTT Technical Research Centre of Finland is a globally networked
multitechnological contract research organization. VTT provides high-end technology
solutions, research and innovation services. We enhance our customers’ competitiveness,
thereby creating prerequisites for society’s sustainable development, employment, and
wellbeing.
Turnover: EUR 300 million
Personnel: 3,200
VTT publications
VTT employees publish their research results in Finnish and foreign scientific journals, trade
periodicals and publication series, in books, in conference papers, in patents and in VTT’s
own publication series. The VTT publication series are VTT Visions, VTT Science, VTT
Technology and VTT Research Highlights. About 100 high-quality scientific and profes-
sional publications are released in these series each year. All the publications are released
in electronic format and most of them also in print.
VTT Visions
This series contains future visions and foresights on technological, societal and business
topics that VTT considers important. It is aimed primarily at decision-makers and experts
in companies and in public administration.
VTT Science
This series showcases VTT’s scientific expertise and features doctoral dissertations and
other peer-reviewed publications. It is aimed primarily at researchers and the scientific
community.
VTT Technology
This series features the outcomes of public research projects, technology and market
reviews, literature reviews, manuals and papers from conferences organised by VTT. It is
aimed at professionals, developers and practical users.
VTT Research Highlights
This series presents summaries of recent research results, solutions and impacts in
selected VTT research areas. Its target group consists of customers, decision-makers and
collaborators.

VTT RESEARCH HIGHLIGHTS 8
Digitising the product process

ISBN 978-951-38-8125-2 (print)
ISBN 978-951-38-8126-9 (online)
VTT Research Highlights 8
ISSN-L 2242-1173
Copyright © VTT 2013
PUBLISHER
VTT Technical Research Centre of Finland
P.O. Box 1000 (Tekniikantie 4 A, Espoo)
FI-02044 VTT, Finland
Tel. +358 20 722 111, fax + 358 20 722 7001
EDITORS: Kaisa Belloni, Olli Ventä
GRAPHIC DESIGN: Tuija Soininen
Printed in Kopijyvä Oy, Kuopio 2013

3
Foreword
At the beginning of the 2010s traditional
heavy industry accounted for 75% of the
total value of Finnish exports, up notably
from 57% in 2000. In addition to industrial
production, the success of Finnish industry is
based strongly on the design and engineering
of devices, working machines, manufacturing
plants, power plants, process machinery and
ships for global markets.
Digitalization has become ever more
vital to the success of industrial produc-
tion and engineering. In practice, it has
enabled the creation and exploitation of a
digital continuum of engineering and opera-
tive computer-based systems. For instance,
requirements analysis and conceptual design
are carried out with the aid of specific com-
puter-based tools. Product and system design
are carried out by a multitude of different soft-
ware tools, often referred to collectively as CAD
(computer-aided design) tools, with each field
of technology and engineering using its own
specific tool sets. Engineering software has
also steadily evolved into extensive product life
cycle management (PLM) systems with CAD
and product data management (PDM) sub-
systems at their core. Engineering is followed
by manufacturing or construction and deliv-
ery, managed again by specific engineering or
manufacturing control software. Thereafter, the
product or system enters its intended indus-
trial use and, depending on the case, effective
operation is governed, for example, by sen-
sor or actuator systems, machine or process
control systems, wider automation, condition
monitoring, or quality management systems.
At a higher level, production planning, enter-
prise resource management (ERP) systems,
and even cross-machine management sys-
tems and networked business management
systems, may be used.
This continuum of engineering and opera-
tive computer-based systems provides the
foundation for the digitalization of industry.
Computers, software and ICT systems have
long been used by all industries. Neverthe-
less, it is more acute than ever to really talk
dually about virtual plants or machines and
real plants or machines. The volume, value
and importance of the digital, virtual realm is
increasing dramatically compared to physical
plants and machines.
Today, everything done in or
for a modern plant is managed
by or with the help of software
and ICT systems. Therefore
being competitive, efficient,
flexible, innovative, experienced,
professional or knowledgeable in
the virtual realm are key competitive
factors for modern industries.
In VTT’s eEngineering spearhead pro-
gramme, ‘Digital product process as a
success factor for technology industries’,
success is defined as having: a) flexible and
comprehensive design and engineering pro-
cesses, b) faster deliveries and engineering
throughput despite the increasing complexity
and size of engineering contents, c) effective
accumulation of engineering knowledge and

4
reuse of proven solutions, d) efficient manage-
ment of growing complexities, engineering
projects, and processes, and f) ascertained
quality. Meeting each of these key criteria to
its fullest extent is an undertaking far beyond
the scope of this research programme and
requiring a vast range of digitalization tools
and systems. The focus of the eEngineering
programme was therefore limited to the fol-
lowing core areas:
1. Simulation-based engineering
2. Knowledge-based engineering, and
3. Interoperability of engineering systems.
The official programme period for eEngi-
neering was from 2009–2012, although some
projects are still on-going in 2013. The pro-
gramme included around 100 projects in total,
the majority of which were funded by VTT or
jointly, for example by the EU or the Finnish
Funding Agency for Technology and Innovation
(Tekes), or partly or wholly by private compa-
nies. The programme had a total budget of
around EUR 40 million, representing about
350 person years. This issue of Research
Highlights focuses mainly on strategic areas of
VTT funding within the programme.
By the beginning of 2009, VTT already
had significant assets in digital engineering
based on its long tradition in leading-edge
modelling and simulation research. The
successful APROS (Advanced Process Simu-
lation Software) environment, for example, is
the culmination of more than 20 years of dedi-
cated research and development in dynamic
process simulation by VTT and its partners.
Based on R&D revenues alone, APROS has
been VTT’s single most valuable ICT software
product. VTT’s BALAS steady-state simula-
tor developed primarily for the pulp and paper
industry has also achieved similar success.
For detailed process simulation VTT has used
and developed a number of CFD (compu-
tational fluid dynamics) packages, and has
also used a wide range of FEM (finite element
method) software for mechanical analyses. In
system-level simulation the Modelica family
has been growing fastest, and VTT is a long-
standing member of the Modelica Association.
VTT has also been active in conducting life
cycle analyses (LCA, carbon footprint calcu-
lations, etc.) and developing LCA software.
For the machine industry, VTT has developed
extensively instrumented virtual studios for
conducting user experience simulations cov-
ering animations, acoustics, thermal comfort,
safety, and ergonomics. Given this back-
ground, the programme implementation was
targeted towards process and mechanical
engineering and, furthermore, to their most
challenging engineering tools, such as model-
ling & simulation and engineering tools co-use.
Product or system platforms are a mod-
ern means of building needed modularity and
efficiency while at the same time enabling
high flexibility and tailorability with respect
to needs. Platforms also provide a means of
accumulating knowledge, managing applica-
tion complexity, and ensuring quality. As the
eEngineering programme demonstrates, the
benefits of platforms can also be realized
in research, by enabling separate research
outcomes, such as methods, algorithms,
analysers, and simulators, to be effectively
combined to produce synergistic applications.
Information exchange across these applica-
tions can be made seamless and live, where
changes in one domain can be effectively
proliferated across tool databases, and where
design items can be made readily avail-
able transparently or via transforms wherever
needed. Single and isolated research tools
may be interesting but greater value can be
obtained from individual research tools by
integrating them into strategic application
platforms. In technical domains, the wider
contexts of systems such as the CAD or
open source tools mentioned above are also
essential and multiply the value of research-
based tools. Platforms are also middleware
systems, meaning that many features that
are necessary for most, if not all, tools need
to be implemented only once as components
of the middleware and be reused elsewhere.

5
Examples include graphical configuration edi-
tors, interfaces to third-party systems, support
for multi-user editing, and security and privacy
features. The use of simulators can be based
on CAD system input or exchange of data,
which is important for increasing user confi-
dence and lowering the threshold for wider
use of simulators.
The most significant achievement of the
programme is Simantics, an extensive oper-
ating system providing an open, high-level
application platform on which different com-
putational tools can be easily integrated to
form a common environment for modelling
and simulation. The development of Siman-
tics began under a previous VTT programme,
Complex Systems Design, which provided the
basic template for the current platform.
A core part of the eEngineering pro-
gramme involved further developing the
template version and directly building the ele-
ments of the Simantics platform. Alongside
this, several existing simulator or tool compo-
nents also needed considerable development
in order to become fully integrated or compat-
ible with the growing Simantics environment.
Other separate simulator and tool develop-
ment projects affiliated with the spearhead
programme also contributed to the evolution
and design of Simantics.
Despite a strong emphasis on the
development of Simantics and its respec-
tive simulator components, the commercial
VirtoolsTM
environment remained the integra-
tion platform of choice for the majority of new
engineering areas for virtual machine cabins,
such as auralization (sound and noise experi-
ences) and thermal comfort. The challenges
related to the platform components and to
our understanding of the interactions involved
led us not to test Simantics compatibility dur-
ing the early stages of the programme. The
Virtools environment also served well as a
means of interfacing with the virtual studio
infrastructure.
In 2013, the advantages and potential
of the programme have become evident, as
summarized in the following:

6
• Process and automation designer’s
working environment extended by a
powerful and versatile simulation environ-
ment, enabled by the Simantics platform.
• Extended workbench supports a wide
range of simulation-based design stages
and purposes, e.g.: a) Early process
dimensioning aided by simple steady-
state simulation, b) Detailed design
iterations by dynamic APROS or even
finer FEM or CFD, c) Early automation
concept testing and later actual automa-
tion system testing against appropriate
level simulators, d) Process simulator
assisted operator training and system
troubleshooting.
• Seamless and transparent bi-directional
exchange of CAD and simulator data,
allowing an iterative working approach
across tools as engineering solutions
evolve. Key CAD software features, such
as version control, externally accessible
to simulators (no need to implement sep-
arately in simulators.)
• Opens a range of possibilities for auto-
matic generation of automation systems
and generation of process details.
• Automation systems can be proven per-
fectly correct by proper methods, such
as formal model checking, and based on
commercially available design tools.
• Combination of design, simulation and
life cycle assessment (LCA) opens new
development possibilities. In addition to
performing traditional analyses and ben-
efitting from existing model libraries and
LCA databases, gaps in LCA data can
be compensated by simulated output
of power plants, manufacturing sites,
etc., and the weaknesses of light LCA
systems compensated by CAD con-
nectivity, advanced user interfaces, and
other strengths of a combined environ-
ment.
eEngineering is built on and has ben-
efitted greatly from previous VTT research
programmes and projects, and this successful
leveraging will continue to be applied in future
programmes. One current example is VTT’s
Multidesign programme, which is aimed at
developing a full chain of simulators spanning
from fine atomic and material grid structures
and the many levels of product structure to
the industrial service business level. The Sim-
antics platform offers a powerful means for
implementing the necessary model interac-
tions and integrations. Another example is the
Smart Grid programme, which is conducting
extensive research on distributed small energy
production by wind turbines, solar systems,
household energy sources, etc., and on
advanced and more accurate energy use by
consumers, vehicles, offices and industry. A
wide range of simulator types are needed to
study the challenges and scenarios presented
by the smart grid concept. Simantics is an
ideal platform for drawing all of these elements
together.
During the eEngineering programme we
had the opportunity to cooperate with many
research and development organizations,
enterprises and professionals in Finland and
abroad. This cooperation has been mutually
inspiring and productive. I wish to thank the
funding organizations, most notably Tekes and
the EU, and also our partner organizations,
most notably the Finnish Metals and Engineer-
ing Competence Cluster (FIMECC). Last but
not least, my sincere thanks to the project
teams and outstanding individuals who have
contributed so much to eEngineering.
Olli Ventä
Programme
Manager

7
Contents
Foreword ................................................................................................................................ 3
Contents ................................................................................................................................. 7
Simantics – an open source platform for modelling and simulation
Simantics – blurring the boundaries of modelling and simulation .................................. 11
Virtual plant combines engineering tools for the process industry ................................ 17
Virtual machines smooth the way from traditional product development
to seamless simulation-based life cycle management ........................................................... 22
Analysing our environmental impact – real and virtual .......................................................... 31
Error-free software through formal methods .......................................................................... 37
Designing user experience for the machine cabin of the future ............................................ 45

8
Authors
Tommi Karhela Pasi Laakso
Research Professor Senior Scientist
Juha Kortelainen Tuomas Helin Antti Pakonen
Principal Scientist Researh Scientist Researh Scientist

9
Simantics — an open source platform
for modelling and simulation

11
Simantics — an open source platform for modelling and simulation
Simulation offers proven advantages as a
tool for modern decision making. In industry,
simulation is widely used in virtual prototyping,
simulation-aided design and testing as well as
in training and R&D. However, obstacles to
wider utilization of modelling and simulation
still remain.
Current modelling and simulation
(M&S) tools exist as separate systems and
are not integrated with other information
management networks. They do not inte-
grate well enough with commonly used
software systems, such as CAD, PLM/
PDM, ERP, or with control systems. Co-use
of the simulation tools themselves is poor,
and the modelling process as a whole is
often considered too laborious.
The Software as a Service (SaaS) and
Open Source business models, used widely in
consumer markets, are also entering the mod-
elling and simulation world. The closed source
licensing model is considered problematic,
especially in public decision making where
the whole computational model should be as
openly available as possible.
To address the boundaries
between modelling and simulation,
VTT developed Simantics, an
integration technique and platform
implementation which has been
published as open source software.
Simantics – blurring the boundaries of
modelling and simulation
AUTHOR: Tommi Karhela
Title: Research Professor
e-mail: tommi.karhela@vtt.ﬁ
Design (CAD) and simulation
system integration
Design systems (CAD) and simulation systems
have traditionally been separate in many areas
of engineering. Notable exceptions include
electronic circuit design and piece goods
manufacturing processes, where simulation-
aided design has long been in use. The reason
for this is also evident. The more deterministic
the target production process is, the easier it
has been to utilize computational models.
In many engineering sectors – such as the
process and construction industries – 2D and
3D CAD systems have already been used for
decades, but these systems do not include
integrated simulation features. Instead, numer-
ous separate computational tools are utilized
in different phases of the engineering process.
Common operating environment
for combining M&S tools
There are many simulation solvers used both in
academia and industry that have sophisticated
computational algorithms, but lack an effec-
tive operating environment. There is a current
need for common operating environments and
pre- and post-processing capabilities as well
as connections to other applications such as
design and control systems.
Pre-processing capabilities include fea-
tures such as 2D-fowsheeting support or
3D-geometry deﬁnition support, discretiza-
tion support (meshing) as well as support for
model validation, model structure browsing
and editing, model component reuse, model
documentation and searching, experiment

12
configuration, model version control and team
features.
Post-processing capabilities include fea-
tures such as 2D-charts and 3D-visualization
of results, 2D and 3D animation of results, and
experiment control visualization. As these are
generic requirements, it is inefficient for dif-
ferent parties to maintain their own individual
operating environments. Instead, a single
common framework could be implemented
which could be jointly maintained and further
developed.
Co-use between different
computational tools
The need for co-use of different simulation
models arises from the same need for design
system integration explained above. The prod-
ucts and production processes modelled are
complex. Heterogeneous multi-level models
are needed which can be utilized across dif-
ferent engineering disciplines. In addition to
Figure 1. The Simantics platform demonstrates its strength in easing communication
between different design and simulation disciplines by enabling smooth data transfer and
information exploitation from design tools to simulation tools and vice versa.
supporting different levels of detail, users also
need to combine optimization and model
uncertainty assessment into their simulation
experiments. In order to break the boundaries
between different computational tools, con-
figuration and simulation run-time integration
is needed.
Team features in M&S
Modelling and simulation environments
are used and developed by extremely het-
erogeneous user populations. Some users
develop new, more efficient solvers and data
structures, others design reusable model
libraries or use these libraries to model real-
world systems, while others simply use these
ready configured models to support decision
making. The team features of a simulation
environment should not support any one user
level alone, they are needed in and across all
of these levels. The features should enable as
efficient as possible reuse of model assets.

13
This requires an infrastructure for publishing
and sharing model components with others
within the same model, project or company, or
even more publicly.
There has been a significant shift in the
software business toward open source and
software as a service business models. For
example, in the US most software startups
that are venture-funded utilize these business
models. The same models are also penetrat-
ing the field of modelling and simulation. It
is likely that future modelling and simula-
tion business will no longer be in platform
solutions, but in simulation components
and services running on open operating
systems for modelling and simulation.
This openness also means that a neutral
democratic forum for decision making on
maintenance and further development has to
be established.
VTT’s Simantics platform has been pub-
lished as an open source environment under
the Eclipse Public License (EPL). To support
democratic decision making, VTT and part-
ners have established a Simantics Division
under the Association of Decentralized
Information Management for Industry
(THTH). At the time of writing, there are 25
company, university and research institute
members in the association.
Main technological innovations of
the Simantics platform
The cornerstone of the Simantics architec-
ture is its open and extensible semantic
data model, which is used to represent the
operating environment and the simulation
and modelling results. The data model is
semi-structured, which means that the data
contains the rules regarding its own struc-
ture. The approach shares similarities with
W3C RDF and OWL, but is especially tailored
for use with engineering and simulation mod-
els. The semi-structured approach allows
the co-existence of different interlinked data
models, which can also be augmented with
new pieces of data when needed. This data-
centric approach places an emphasis on high
quality representation, which increases the
usefulness of the produced results. Simantics
comes with standard models developed for
common simulation and modelling patterns.
As an example, a generic conceptualization
of hierarchical, connected and parameterized
models can be used as a basis for differ-
ent domain models. The data models are
organized in layered conceptualizations, i.e.
ontologies, which can be developed before
or during modelling.
As an operating system for modelling
and simulation, Simantics needs to be able to
handle various data models related to different
tools, computational methods and modelling
methodologies. Successful co-use of these
different domain services requires power-
ful integration and mapping tools between
the different models. Simantics addresses
these needs by supplying an ontology-based
mapping framework for mapping and trans-
Whereas modelling and simulation is
used widely in design and development,
it has yet to gain a foothold in operational
decision making. In this area, simulation
models are connected to measurements
from real systems and predictions
are made to support actual decisions.
Co-use of simulators and control
systems sets requirements for the real-
time communication, synchronization
and simulation control facilities of
the integration platform. If these are
implemented in a neutral and efficient
way, they can also provide a solid base for
the communication and synchronization
of different dynamic simulation tools
or for the high-level parallelization of
several simulation experiments. High-level
parallelization here refers to process-
level parallelization i.e. several simulator
instances running in parallel, not to code-
level parallelization of a single simulator.
High-level parallelization is also useful in
optimization and uncertainty quantification
assessment cases.

14
forming models within Simantics. The general
approach is to import domain models into
Simantics as they are, and then transform the
data further by using semantic mechanisms,
which have been studied for some time [e.g.
1, 2, 3]. The advantage of this approach is
that each specific data model is kept clean
and separate. Elaborate data transformation
mechanisms are also useful for generating
models from other models. From the use case
point of view, the mechanisms for mapping
and transformation enable the co-use of dif-
ferent domain models as well as the co-use of
models of different levels of detail. Simantics
offers the user a special Simantics Constraint
Language (SCL) for developing user config-
urable mappings and transformations. The
same functional programming language can
be used, for example, for semantic queries
and model validation.
The use cases for modelling needs in a
data-driven simulation and modelling sys-
tem are highly versatile. In addition, the
semantic data modelling approach is heavy
performance-wise. To be able to fully establish
a semantic data driven modelling approach,
Simantics supports a wide range of mecha-
nisms for extending the application range of
the semantic data model. Simantics offers
seamless support for four levels of persistence
of semantic data in a unified model. Memory
persistent parts of semantic data can be used
to model quickly changing and transient struc-
tures, so that the structures only exist during
a modelling session. Workspace persistent
structures are only stored in the user’s local
hard disk and can be used to represent vari-
ous cache or preference structures, which are
generated or otherwise not publishable to all
users of the distributed database. Database
persistent parts of the data model are shared
by different clients of a database server. Data-
base persistent data is also fully versioned.
Finally, database persistent data can also be
published and synchronized between data-
base servers, called team servers, across
organizations.
The wide range of persistence levels
and representations is common in simulation
cases. For a semantically similar attribute we
can have input values in engineering systems,
permanent configuration values in simula-
tion models, different sets of, for example,
dimensioning values in simulation models,
computed result values and time series, real-
time dynamic simulation or measurement
values, etc. The representation of values for
the same attribute can be modelled in com-
pletely different ways or not at all. Some
attributes have many values, some are time-
dependent, some are persistently stored and
some are not. To fulfil the integration goals
the system needs to be able to represent
and manage all these different pieces of data
and, most importantly, to associate the data
semantically together so that the data can be
integrated. Simantics addresses these issues
through semantic modelling of variables and
their values and simulation experiments and
by specifying a software interface to be used
to obtain values for a given configuration of
semantically modelled variables. The inter-
face defines a semantic connection from a
data value to the concepts of the data model
while allowing free acquisition of values from
any source. In many cases the obtained val-
ues are backed by the semantic data model,
but can also be directly obtained from, for
example, a simulator or measurement device.
This framework for simulation data manage-
ment makes Simantics unique among other
data modelling platforms.
Simantics – an open operating
system for M&S
The benefits to industry of modelling and simu-
lation are clearly proven, yet two key obstacles
to the use of simulation still persist – cost
and timely availability of simulation models.
These are both the result of model develop-
ment not being integrated into engineering
work flow and data management. Recent
cases have shown that a sufficient system-
level model can be created based entirely

15
on engineering data. These case studies are
explained in more detail in the next chapter. It
has been additionally concluded that the co-
use of different modelling and simulation tools
is currently insufficient. Multi-scale models
combining different levels of detail would ben-
efit from better configuration integration and
run-time co-use of different simulators. Fur-
thermore, the co-use capabilities of simulation
environments and real-time systems, such as
control systems, are inadequate. The chal-
lenge of integrating design system features,
simulation features and real-time control and
measurement features into the same software
architecture has been identified. Current simu-
lation environments also lack team features,
which are essential in modern globally net-
worked engineering projects.
VTT has introduced a solution that uti-
lizes a semantic data modelling approach
and combines this expressively powerful
ontology-based design with fast acces-
sibility to simulation, measurement and
control data. This data-driven approach
opens possibilities for automatic model
validation, reporting, processing, anno-
tation and linking. The layered ontology
structure enables expandability and reusabil-
ity. The heart of the integration solution is an
ontology-based mapping mechanism that
enables rule-based synchronization of differ-
ent engineering and simulation models. The
idea is to integrate data from different
background systems into the environment
‘as is’ using native data models. The model
mapping is done inside the platform using
ontology-based mapping rules. To address
problems of scalability and processing speed
– key bottlenecks in the semantic approach
– the developed solution has been optimized
for industrial use. The platform has been pub-
lished under an open source license and is
maintained jointly with industry partners in the
form of an association.
Figure 2. Evolution of information management.

16
References
[1] Maedche, A., Motik, B., Silva, N. & Volz, R.
2002. MAFRA - A Mapping FRAmework
for Distributed Ontologies. Proceedings
of the 13th International Conference on
Knowledge Engineering and Knowledge
Management, Siguenza, Spain. Pp. 235–
250.
[2] Pierra, G. 2004. The PLIB Ontology-Based
Approach to Data Integration. Proceedings
of the IFIP 18th World Computer Congress,
Toulouse, France. Pp. 13–18.
[3] Qian, P. & Zhang, S. 2006. Ontology
Mapping Meta-model Based on Set and
Relation Theory. Proceedings of the First
International Multi-Symposiums on Com-
puter and Computational Sciences Volume
1 (IMSCCS’06), Hangzhou, China. Pp.
503–509.
Related publications
Karhela, T., Villberg, A. & Niemistö, H. 2012.
Open Ontology-based Integration Platform
for Modeling and Simulation in Engineering.
International Journal of Modeling Simulation
and Scientiﬁc Computing, Volume 3, Issues 2:
(1250004), World Scientiﬁc Press.

17
Process plant deliveries consist nowadays
of two plant components – a ‘real’ plant and
a ‘virtual’ plant. The real plant is the actual
nuts and bolts delivery, while the virtual
plant comprises all of the digital material
handed over to the customer. However, a
lack of proper standards and harmonized
procedures regarding the content and
delivery of virtual plants poses a number
of challenges for system integrators (EPC
contractors) – to the extent, in fact, that vir-
tual delivery may be more challenging than
the actual plant delivery itself.
A live or executable virtual plant
combines a virtual plant, i.e. piping and
instrumentation (P&I) diagrams, 3D models,
automation models and electrical designs,
together with simulation models and other
computational algorithms. The live virtual
plant has a wide range of potential uses
throughout the life cycle of the facility, from
the early design phases through to commis-
sioning and operation support. Some of the
most time-consuming tasks involved in live
virtual modelling include the collection of
the initial data needed to create simulation
models and keeping this data up to date
throughout the life cycle of the plant. There
have previously been no readymade work
processes, standards or tools for efficient
simulation model generation. To address
this problem, VTT has developed infor-
mation model integration techniques that
enable seamless bi-directional data flow
between engineering systems and simula-
tion models, facilitating the development
Virtual plant combines engineering tools for
the process industry
AUTHOR: Pasi Laakso
Title: Senior Scientist
e-mail: pasi.laakso@vtt.ﬁ
of efficient working processes. This article
gives an overview of two key development
efforts in this area.
VTT has been developing simulation tools
for the process industry for well over 25 years.
Apros – a software platform for dynamic mod-
elling and simulation – has been one of the
biggest success stories in this area [1]. Apros
has been used for simulation-assisted auto-
mation testing [2], as a tool for automation
modernization of the Loviisa nuclear power
plant (see Figure 1), and in control concept
development [3]. One of the main obstacles
to wider utilization of these beneﬁcial methods
has been the amount of work needed to cre-
ate and update simulation models. Now, there
are new possibilities to solve the problem.
Accurate and up-to-date virtual plant
models are becoming more common. EPC
contractors have started to integrate EDMS
(Engineering Data Management System) tools
such as Comos or Intergraph SmartPlant
as a part of their normal design procedure.
Industry is also adopting more and more so-
called intelligent CAD systems that provide
online access to up-to-date databases that
always contain the latest engineering data.
Meanwhile, VTT has been developing the
Simantics integration platform [4], which
can be used to connect databases and
simulators together, providing the basis for
creating live virtual plants.
New techniques make it possible to
take simulation into use earlier and use more
detailed simulation models cost-effectively.
Updating design data needed for simulation
OTHER CONTRIBUTING AUTHORS:
Jari Lappalainen, Tommi Karhela, Marko
Luukkainen

18
Figure 1. Simulation is widely used in the Loviisa nuclear power plant automation
renewal, e.g. to develop and test the new automation system and for operator training.
can be done automatically several times dur-
ing the design process and using real design
data every time.
Recent work in this area has focused on
automatically generating simulation models
based on design data, and then transfer-
ring possible changes back to the plant
engineering databases. The possibility of
simulation-assisted preliminary planning has
also been considered. In the latter case, the
plant design database would be seeded with a
design created in a simulator. A typical integra-
tion project would include a technical solution
for transferring design data and, more impor-
tantly, finding correspondences between the
data objects of the simulation model and the
plant design database.
Case: Foster Wheeler Energia Oy –
connecting Apros to Comos
Foster Wheeler Energia Oy (FWE) has been
using Apros simulation software for modelling
boilers and related automation. They have also
integrated Comos [5] as a part of their boiler
engineering process. VTT has been involved
in the take-up process and has integrated
dynamic process simulation as part of the
engineering process by designing and imple-
menting tools for integrating Comos and the
Apros simulation tool. The developed tools
transfer process and automation design data
from the Comos database to the Apros simu-
lator and generate or update the Apros model
automatically.
The project began by analysing FWE’s
existing simulation practices. Based on this,
the integration tool requirements were iden-
tified. Three potential levels of detail of the
dynamic simulation model were considered.
The first level – the conceptual level – con-
tains the main process components and the
pipelines between them and is usually tuned
to match steady-state performance calcula-
tion. Simulation at this level was considered
most beneficial in the FWE case and also
in future cases in general. The second level
of detail, referred to as the basic level, is
more detailed, corresponding to the level of

19
detail of a piping and instrumentation (P&I)
diagram, with most pipes and similar heat
surface groups described individually, and
also including detailed automation diagrams.
Simulation at this level is highly extensive
and detailed, and has previously not been
considered cost effective. The third level of
simulation considered was the 3D level, in
which even the smallest devices and pipes
are modelled. This level of detail was not
implemented.
Figure 2 shows the typical corre-
spondence between a P&I diagram and a
conceptual-level model. The P&I diagram on
the left shows an economizer system and
its header and pipe connections to other
parts of the plant. Two pipe bundles are
also described. The diagram shows only the
water side (flue gas side is not shown). The
corresponding Apros model is shown on the
right. In the Apros model, the input header
represents all input headers and pipes in the
system combined and, likewise, the output
header represents the output headers and
pipes combined. Similarly, the pipe bundles
identified in the P&I diagram are modelled as
a single heat exchanger in the Apros model.
The identification of system components is
based on the KKS power plant classification
system.
A key challenge brought to light by the
FWE project was a lack of readily available data
for simulation model generation. This under-
lines the importance of planning simulation
tools and EDMS tools in parallel. For exam-
ple, obtaining accurate elevation levels during
the early design stages was problematic, as
elevations are typically defined only later in the
project during the detailed 3D design stage.
It was also found that not all plant modelling
practices were a suitable basis for simula-
tion; for example, a P&I diagram might appear
fully connected visually, but contain symbols
that are not present in the plant model. When
generating a simulation model, this results
connections being missed. This shows that, in
addition to appropriate and timely selection of
tools, also further guidance of plant designers
is needed.
‘Dynamic simulation with Apros has been
valuable tool for us when investigating
boiler plant configurations. Verifying and
optimizing plant design decisions is an
important part of the work process. When
we started to integrate Comos as part of
Figure 2. Economizer P&I diagram and corresponding conceptual-level simulation model.
The heat surfaces, i.e. pipes, in the middle generate heat exchanger in Apros. Headers
and pipes entering and leaving the system are combined together as heat pipes in Apros.

20
our design process, it was natural to also
include connection to Apros as part of the
Comos development.
Currently, we can develop the plant with
our design tools, run steady-state analysis
using our in-house dimensioning tool and
then automatically generate Apros model
for use in analyses requiring a dynamic sim-
ulation model. Connection to our own boiler
model can also be generated automatically.
If the design changes, the model can be
updated or regenerated easily. Now we can
virtually test more design alternatives in a
shorter time than before.’
-Jenö Kovacs, D.Sc.,
Principal Research Engineer,
Foster Wheeler Energia Oy
Case: Fortum – connecting Apros
to SmartPlant
In another on-going project, the Apros simula-
tion tool is being integrated with Intergraph’s
SmartPlant product family [6], particularly
SmartPlant P&ID, SmartPlant Instrumenta-
tion, and SmartPlant Foundation. The project
shares many similar features to the previous
FWE case concerning data transfer between
process modelling and P&I diagrams. The
project’s prime focus, however, is on simula-
tion-based basic automation design and its
integration with other SmartPlant engineering
tools. Data transfer with SmartPlant is achieved
through SmartPlant Foundation, which ena-
bles electronic management of all of the plant’s
engineering information, integrating data on
physical assets, processes, and regulatory and
safety imperatives. Apros is used as an automa-
tion design tool, while SmartPlant Foundation
acts as an integration platform between auto-
mation, process and instrumentation planning.
Typical automation design solutions and struc-
tural automation components are accepted
and taken into use by other developers through
SmartPlant Foundation.
The integrated solution enhances the
use of SmartPlant products by adding con-
trol design and Apros-based testing features
to it. The integration also benefits SmartPlant
owner operators by extending access to engi-
neering asset data, including dynamic plant
performance versus as-designed engineering
data. The solution also enhances the use of
Apros by enabling use of SmartPlant engi-
neering data to create dynamic models more
efficiently.
‘Fortum considers dynamic simulation as
an essential tool in modern power engi-
neering and foresees its role as clearly
increasing in the future. Real breakthroughs
can be achieved through seamless inte-
gration between engineering project tools
and simulation software. Besides technical
integration capabilities, also a willingness
to move towards new working methods is
needed. Further development is needed,
but we see great potential here. Embracing
this approach could bring a competitive
edge to the Finnish engineering sector.
Fortum wants to be on the front line of this
development, firstly as a company with a
strong engineering tradition, and secondly,
as a committed simulation provider for the
power industry.’
-Sami Tuuri, Product Manager,
Fortum
Discussion
Dynamic simulation serves as a valuable tool
in plant design and modernization, enabling,
for example, automation and process design
to be verified ahead of plant construction, and
plant personnel to be effectively trained to
operate the plant under normal and abnormal
conditions.
This development provides a good basis
for wider take-up of simulation in industry.
There is still work to do. The FWE/Comos
integration project has progressed to the
maintenance phase. The system has been
successfully tested with completed engineer-
ing projects, but with the emergence of new
projects alterations to the current generation

21
The Simantics integration platform
developed by VTT has been
successfully used (see case Foster
Wheeler above) to integrate the
Apros simulation tool with the
plant asset management software
Comos and, in an ongoing Fortum
collaboration, the platform is also
being used to integrate Apros
with SmartPlant Foundation. An
integration between Aucoplan and
Apros has also been developed in
an earlier case [3]. Automatically
generated models are free of manual
copying errors, faster to generate
and can be more detailed than
normally possible.
rules are likely to be needed. Apros-Comos
integrations also need to keep pace with the
evolution of Comos. Deeper integration with
SmartPlant is also currently under develop-
ment ahead of the piloting phase. It should
also be noted that as Comos and SmartPlant
are intended to be used ﬂexibly in different
environments, a degree of tailoring is needed
when applying the technology within new
companies.
References
[1] Apros web pages www.apros.ﬁ
[2] Tahvonen, T., Laakso P., Wittig, J.,
Hammerich, K. & Maikkola, E. 2009.
Simulation Assisted Automation Test-
ing During Loviisa Automation Renewal
Project., 6th IFAC Symposium on Power
Plants and Power Systems Control, 5–8
July 2009, Tampere, Finland. Power
Plants and Power Systems Control, Vol-
ume 1 | Part 1., Pp. 314–319.
[3] Paljakka, M., Talsi, J. & Olia, H., 2009.
Experiences on the integration of auto-
mation CAE and process simulation tools
- case Fupros. Automaatio XVIII Semi-
naari 17.-18.3.2009, SAS, julkaisusarja
36. Finnish Society of Automation. Hel-
sinki.
[4] Simantics web pages www.simantics.org
[5] Comos web pages http://www.
automation.siemens.com/mcms/plant-
engineering-software/en/Pages/Default.
aspx
[6] SmartPlant Foundation web pages http://
www.intergraph.com/products/ppm/
smartplant/foundation/default.aspx

22
The use of computational tools in product
development is now standard practice in
mechanical engineering and design. The
development of a modern, complex high-
technology product, such as a modern
passenger car, aeroplane, or diesel engine,
would be practically impossible without
computer-aided design (CAD) systems, com-
putational analyses, and system simulation.
These provide the tools for designers to gain
valuable feedback about the behaviour and
performance of products under development.
Rapid advances in computer technology and
decreasing computational costs have made
it possible for even small companies to incor-
porate digital design and simulation into
their product develop-
ment. At the same
time, the most
a d v a n c e d
c o m p a n i e s
are tak-
ing steps
towards the
vision of fully
digital prod-
uct processes
and digital man-
agement of all
data, information, and
even knowledge related to their product pro-
cesses. The vision includes digital functional
product models that can be used to virtually
test products and their behaviour from multi-
Virtual machines smooth the way from
traditional product development to
seamless simulation-based life cycle
management
AUTHOR: Juha Kortelainen
Title: Principal Scientist
e-mail: juha.kortelainen@vtt.fi
ple engineering perspectives as well as data
management solutions that enable engineers,
designers, and other doers in the process to
efficiently utilize all of the information involved
in the product process.
The mechanical engineering research
carried out under the eEngineering research
programme focused on the concept of a sim-
ulation-based product process covering the
entire product life cycle and on the integration
of data and engineering software applications
into this process. One of the major findings
of the programme was the importance of
product information and related knowledge,
and especially how data, information and
knowledge are managed and in what form.
Computational tools and systems often store
data in forms that are not known or under-
stood by their users, and not always even
by the system managers – data simply goes
into a database, and there it stays. However,
together with engineering knowledge, the
information contained in this data is essen-
tial capital in the product process. This
capital is often not systematically man-
aged at all. The technologies developed
in the eEngineering programme, such as
the Simantics platform, provide means
and practical tools for managing the valu-
able information and knowledge of the
product process. Another important insight
was the need to expand the application of
simulation beyond estimating the behaviour
of physical systems to include non-physical
As the offering of
different computational
tools, design systems and
data management solutions
continues to surge, the need to
separate data, information and
knowledge from computational
tools is becoming
increasingly evident.

23
systems and processes, such
as the function of an organi-
zation or the markets. Even
though these simulations
may not be perfectly
accurate or fail-safe, they
provide a similar learning
experience as the simu-
lation of physical systems
provides to an engineer or
designer. The modelling phase
of the target system helps the user
to understand the structure, relations, and
scale of the components and phenomena
involved in the system. By simulating the
model, the user gains understanding of the
behaviour and dynamics of the system and
the interaction of its components and sub-
systems. This applies to both physical and
non-physical systems, and helps the user to
design better products, processes and ser-
vices.
From product
development to
simulation-driven life
cycle process
VTT has an excellent van-
tage position for viewing the
landscape of Finnish and
international industry. As a
neutral technology devel-
oper and know-how producer,
VTT has a comprehensive under-
standing of the use of computational
methods and tools in industrial product pro-
cesses. The key challenge is strong case or
conditions dependency. What is obvious to
one industrial sector or company does not
necessarily apply to others. We view a product
process as a whole chain, including tools and
systems, people and practices, the operating
environment and the markets. A product pro-
cess is not just about technology, but about
everything related to the product life cycle.
The concept of
applying simulation for
estimating the behaviour
of physical systems can be
extended to non-physical
systems and processes,
such as the function of
the organization or the
markets.

24
Development levels in the simulation-based product process
On the ﬁrst level, simulation in product development, computational methods are used to study detailed
and local phenomena (from the product perspective). At this level, computational methods typically
have only a limited effect on product development, and most product development is carried out using
traditional methods, i.e., engineering practices and established design principles. This is the prevailing
practice in many mechanical engineering companies in Finland today.
On the second level, virtual prototypes in product development, the whole product or at least its
main subsystems are modelled and simulated as virtual prototypes. This enables engineers to study
the overall dynamics and behaviour of the system and gain important information at an early stage in
the product process, before constructing any physical prototypes. Product development is still based
on traditional engineering practices. Some companies in the mechanical engineering sector in Finland
utilise this approach.
On the third level, simulation-based product development, similar computational methods are used
as on the previous level, but now they are used systematically and form the basis of the design process.
Before design work commences, the component, subsystem or product in question is modelled and
simulated using a coarse model to gain understanding of the interactions and interferences involved.
Based on this knowledge the design is then improved, and the procedure is continued iteratively until
a design that fulﬁls the requirements is achieved. The challenge of this approach is in implementing this
design process in practice. Only a few mechanical engineering companies in Finland operate on this
level.
On the fourth level, simulation-based product process, the product modelling and simulation con-
cept is also applied to the product process. This means that in addition to modelling the physical
properties of the product, also the non-physical processes, such as the product maintenance business
model and product development organization, are modelled and simulated. In addition, other important
aspects, such as carbon and water footprint and other environmental effects, are analysed based on
the available product data, and the life cycle performance of the product is optimized. This is still the
future, although methods and tools to implement this level are already available.
VTT provides research and services in all key areas of the simulation-based product process con-
cept.
Figure 1. Evolution of simulation use in the product process and the increasing importance of data
management. [2]

25
VTT’s long experience in applying com-
putational methods and simulation in research
and in industrial product development has pro-
vided a unique understanding of how the use
and development of computational methods
evolve within an organization. This evolution
process can be roughly divided into four lev-
els, as illustrated in Figure 1. The objective
of modelling and simulation is to gain better
understanding of the systems in the product
process and, ultimately, to create a ‘big pic-
ture’ of not only the product, but the whole
product life cycle. The different simulation
methods and tools, optimization algorithms,
and data integration solutions available are the
pieces in this puzzle, which need to be col-
lected, modified and combined in order for the
big picture to be revealed.
The concept of the simulation-based
product process, i.e., the use of computa-
tional methods for simulating and analysing
the whole product life cycle, and different
approaches to software integration through
a centralized data management system,
were studied together with the Finnish Fund-
ing Agency for Technology and Innovation
(Tekes) in the joint-funded research project
Computational models in product life cycle
– Codes [1]. The project shed new light on
the concept of simulation-based product
life cycle management and the fact that the
same justifications for utilizing simulation in
studies of physical systems are also valid for
larger-scale systems and processes. Another
important finding was that we already have
many of the pieces of the ‘big picture’ of
whole life cycle simulation, and we have a
good understanding of the parts needed to
complete the rest of the puzzle. The project
was carried out under VTT’s eEngineering
programme as part of the national Digital
Product Process research programme.
Simulation-based design requires
strong data management
As product processes become increasingly
comprehensive, covering all process aspects
and business perspectives, effective product
life cycle management at the product devel-
opment and manufacturing stages is also
becoming ever more important. Product
data, information and related knowledge
are the most valuable assets of the prod-
uct process. The software applications and
systems used to produce, modify and store
data during the process have a direct impact
on the functioning of the process. For exam-
ple, the engineering software applications
used in product and manufacturing design
influence the working efficiency of the design
engineers and, in turn, the end quality of the
design. This is emphasized in the simulation-
based product process, where the design of
the product and the manufacturing process
are based heavily on the results provided
by the system simulation tools used and on
the understanding gained from them. Also,
as simulation and design are carried out
iteratively, the amount of data gathered is
extensive, and effective data exchange and
data integration become essential.
Even though digital design and com-
putational methods are widely used in
mechanical engineering, design and simu-
lation data exchange is still cumbersome in
daily research and development work. Dif-
ferent design systems and simulation tools
use their own data formats and internal
data models, which are not well-supported
by other software applications. Switching
Computational methods provide
a valuable tool for estimating the
effects of design and strategy
decisions in the early stages of
the product process and thus
enable overall optimization of the
product and its related services.
This requires strong solutions
for product and design data
management, especially in globally
operating organizations.

26
CAD systems mid-project can
be a formidable undertaking
due to data exchange issues
and should preferably be
avoided. Exchanging data
between simulation soft-
ware is in some cases totally
impossible due to a lack of
common data formats. This
means that the software appli-
cations used in the process have a
strong effect on the daily work of the design
engineer, and the choice of tools used deter-
mines many outcomes later in the process.
As an example, if structural analysis model-
ling has been started with one commercial
FEM software application, it can be very dif-
ﬁcult to switch later in the process to using a
different, preferred FEM application for com-
putations. In an ideal situation, modelling
data would be independent of the com-
putational tools used, and tool selection
would be freely possible at the computa-
tional analysis stage.
The general concept of
separating product and design
data from the tools used
to produce, modify, and
process it is illustrated in
Figure 2. The concept also
includes the management
of engineering knowledge in
machine-understandable form.
According to the concept, all data,
information and knowledge related to
the product process are managed and inter-
preted by a centralized system. This ensures
that product information is preserved even if
the computational tools used are changed. In
addition, the concept provides the freedom to
choose the right tool for each purpose, thereby
enabling the use of computational resources
to be optimized. The centralized data man-
agement approach allows all users to access
up-to-date data, whether the organization is
local or worldwide.
The vision of separating product data
and computational tools, described above
Separating the
product data and
the computational tools
and systems ensures the
preservation of product
information and provides
the freedom to choose
the best computational
tools.
Figure 2. The concept of separating valuable computational product data from the tools
that use it.

27
and illustrated in Figure 2, requires a range
of skills, systems and methods. Each com-
putational method requires specific expertise
to be gathered in order to reliably model and
simulate complex physical systems. Running
large computational analyses also requires
large computational resources and systems
capable of utilizing these resources. The vast
amounts of numerical data produced by large-
scale simulations need to be managed and
used efficiently in the product process. These
are considerable challenges in themselves and
represent a major combined undertaking, but
the benefits are compelling. The Simantics
platform introduced in the previous article
offers a solid basis for achieving these data
management goals.
Case: Data architecture and
software application integration
through a centralized data
management system
The eEngineering programme studied and
implemented the ‘big picture’, discussed in
the previous sections, in a number of areas.
A general architecture for data and software
application integration into a centralized data
management system, such as the Simantics
platform, was designed, and selected soft-
ware applications and formats were integrated
to determine the effort required for its imple-
mentation and to understand the concrete
process and its details. The data model and
software application integration architecture is
illustrated in Figure 3. The main design idea is
that the data model of a software application
or a file format, such as the Abaqus INPUT
format, is created one-to-one in the Simantics
platform as an ontology. The data from differ-
ent software applications is integrated inside
the Simantics platform using semantic data
mapping based on general software appli-
cation or domain ontologies (e.g., a generic
FEM ontology). This approach simplifies
the implementation of laborious software
application integrations and enables the
utilization of data mapping features of the
semantic data representation approach.
Figure 3. Architecture of the integration of the simulation software applications into the
common data management solution.

28
Several data models, such as the Uni-
versal File Format (UFF) and the Abaqus
FEM package’s INPUT format, were created
as ontologies in the Simantics platform,
and the necessary file parsers were imple-
mented. Based on new knowledge of the
required effort and the process, the use of a
parser generator (ANTLR3) was studied and
a demonstrator implementation of a data
parser was created. The conclusions of the
work on data modelling and integration are:
• The concept of integrating engineering
software applications according to the
architecture presented in Figure 3 is
valid. Data transfer from one software
application to another is straightfor-
ward and software integration into the
data management system is lossless,
due to explicitly matching data types
and structures in both systems (i.e., the
engineering software application and
the data management system). Data
mapping from a software application
specific ontology to another ontology
is easy compared to mapping between
separate software applications. This is
due to the built-in data mapping mecha-
nisms in semantic data representation.
• High-level software development tools,
such as parser generators, can mark-
edly improve the implementation of
software and data integration and
ease software maintenance and further
development of the component.
• The amount of work needed for
implementing all necessary software
application integrations for the needs
of a whole product process is large and
cannot be carried out by one actor in
a software or engineering eco-system.
Thus, communication, architecture
design, and standardization between
data models and system interfaces are
needed. This is an opportunity for soft-
ware vendors and service providers to
find new business areas.
Case: Implementing a
demonstrator of a multibody
system simulation environment
The advantage of the platform concept was
highlighted during the execution of the eEngi-
neering programme in a number of key ways.
Platforms are a natural continuation of the
vision of simulation-based product process
and life cycle management discussed in the
above sections, and the concept of separat-
ing valuable product data and computational
engineering tools, as illustrated in Figure 1.
Practical examples concretizing the use
and usefulness of the platform concept were
studied and demonstrated in the eEngineering
programme. The main features of the Sim-
antics platform are described in brief below
(more detailed description of the platform is
presented in chapter Simantics – blurring the
boundaries of modelling and simulation). The
Simantics platform consists of a background
data management system (a semantic data-
base), a graphical user interface framework,
and high-level software components on the
framework, such as a 2D model graph editor
and model structure browser. The availabil-
ity of these high-level components markedly
improves the efficiency of software implemen-
tation and also enables quick proof of concept
testing of new ideas for software applications.
In mechanical engineering, this was tested
by implementing a 3D solid geometry model-
ler component on the Simantics platform and
using the component to create a 3D multi-
body system (MBS) model editor, integrating
an existing simulation environment as the
computational backend, and implementing
post-processing features, including simulation
results animation and a plotting component
(see Figure 4). For numerical solving of the
MBS simulation, another platform, the Open-
Modelica Environment1
, was used. For the
implementation, external high-level software
components, such as the OpenCASCADE2
geometry kernel for the 3D solid modelling,
and the Visualization Toolkit (VTK)3
for 3D
graphics, were used to speed up the software
1
OpenModelica project: www.openmodelica.org
2
OpenCASCADE project: www.opencascade.org
3
Visualization Toolkit software: www.vtk.org

29
implementation. The OpenModelica Environ-
ment is an open source implementation of the
Modelica4
simulation language environment.
Implementing the 3D solid geometry model-
ler and the MBS modelling, simulation, and
post-processing features on the Simantics
platform required a total labour input of four
man-months.
Based on the case study, the following
conclusions were drawn regarding the applica-
tion of platforms to build a data management
system for a simulation-based product pro-
cess and to improve software development
efficiency:
• Centralized, open, and extensible data
management systems are needed to
implement the vision of a simulation-
based product process and product life
cycle process. A good proof of concept is
the Simantics platform.
• The platform concept was studied, tested
and proofed in mechanical engineering by
implementing a 3D solid modeller and a
3D MBS model editor, simulation man-
ager, and post-processing features within
four man-months. This was enabled by
the existing high-level software compo-
nents and platforms.
Bringing the vision to reality
The old maxim ‘knowledge is power’ has a
new meaning in the context of product life
cycle data management, but when it comes
to the products and services in this business
area, it has deep wisdom. The one who man-
ages the data, information and knowledge
in the product process, and can utilize it effi-
ciently, has a clear advantage in the market.
From the research point of view, the key ques-
tion is what steps and actions must be taken to
ensure that Finnish companies are at the fore-
front of this development and reap its benefits?
Figure 4. The graphical user interface of the MBS editor built on the Simantics platform,
showing a test model.
4
Modelica simulation language project:
www.modelica.org

30
The eEngineering programme has dem-
onstrated the importance of data, information
and knowledge management in the product
process. The case studies and the lessons
learned from them have provided understand-
ing of the size of the challenge, the ways to
implement the vision, and the importance of
long-term work in the form of well-designed
technologies and platforms. Implementa-
tion of the vision requires close cooperation
between different parties: research, commer-
cializers, service providers, and end users.
The challenge is big, and must therefore be
divided into manageable pieces, e.g., using
standardization and open specification of
required system and data interfaces, and
several parties are needed to implement it.
The development trend among vendors of
big design systems is clearly towards a simi-
lar vision. End users who adopt the concept
and the vision of the pioneers will reap the
benefits. Players who wait for proof of con-
cept from the market will most likely be too
late.
References
[1] Kortelainen, J. 2011. Overview to the
Codes Project. Espoo: VTT. 24 p. (VTT
Research Report VTT-R-03753-11.) http://
www.vtt.fi/inf/julkaisut/muut/2011/VTT-R-
03753-11.pdf.
[2] Glotzer, S., Kim, S., Cummings, P., Desh-
mukh, A., Head-Gordon, M., Karniadakis,
G., Petzold, L., Sagui, C., & Shinozuka, M.
2009. International assessment of research
and development in simulation-based engi-
neering and science. WTEC panel report,
World Technology Evaluation Center, Inc.
WTEC. 396 p. http://www.wtec.org/sbes/
SBES-GlobalFinalReport.pdf, (accessed
November 7, 2012).
Benioff, M. & Lazowska, E. 2005. Com-
putational science: Ensuring America’s
competitiveness. Report, President’s Infor-
mation Technology Advisory Committee
(PITAC),. 104 p. http://www.nitrd.gov/pitac/
reports/20050609_computational/computa-
tional.pdf, (accessed November 7, 2012).
Kortelainen, J. 2011. Semantic Data Model
for Multibody System Modelling. Espoo: VTT.
119 p. + app. 34 p. (VTT Publications 766.)
ISBN 978-951-38-7742-2 (printed), 978-951-
38-7743-9 (online). http://www.vtt.fi/inf/pdf/
publications/2011/P766.pdf.
Kortelainen, J. & Mikkola, A. 2010. Semantic
Data Model in Multibody System Simulation.
Proceedings of the Institution of Mechani-
cal Engineers, Part K: Journal of Multi-body
Dynamics, Prof Eng Publishing, Vol. 224, No.
No.4, pp. 341–352.
Oden, J., Belytschko, T., Fish, J., Hughes,
T., Johnson, C., Keyes, D., Laub, A.,
Petzold, L., Srolovitz, D., Yip, S., & Bass, J.
2005. Simulation-based engineering science.
Report, National Science Foundation. 66 p.
http://www.nsf.gov/pubs/reports/sbes_final_
report.pdf, (accessed November 7, 2012).

31
Life-cycle assessment — Life cycle
inventory — Impact assessments
— Software tools
Rising global population and material well-
being present a multitude of global challenges,
including biodiversity loss, climate change,
ocean acidification and competition over
scarce material and fossil energy resources
such as phosphorus and crude oil [1]. To
mitigate these impacts, we need to transform
our societies, business concepts and indus-
trial processes towards high energy- and
resource-efficiency and reduced environmen-
tal pressure. Modelling and comparing the life
cycle environmental impacts of different sys-
tems enables us to identify the hotspots in
value chains with most improvement potential
and enables selection of production path-
ways with smallest environmental impacts. A
focus on the whole life cycle instead of partial
optimization of single steps in the value chain
ensures that no burden shifting through par-
tial optimization takes place. VTT provides the
know-how, technology and software tools for
modelling and improving the environmental
performance of our society and production
systems, in parallel with economic and techni-
cal considerations.
Typically, engineering design is done using
CAD systems and environmental assessment
using LCA software. While these two worlds
are strongly related – especially in ecodesign
– they are not necessarily connected. CAD
systems contain a wealth of information about
different parameters of the product life cycle
from which Life Cycle Assessment (LCA) can
Analysing our environmental impact – real
and virtual
AUTHOR: Tuomas Helin
Title: Research Scientist
e-mail: tuomas.helin@vtt.fi
benefit. Stronger integration between CAD
and LCA software system has been sug-
gested by several researchers, for example, in
[2] and [3]. In the eEngineering spearhead
programme we have implemented bridges
between Simantics and Intergraph Smart-
Plant Foundation (CAD integration system)
and Siemens Comos (CAD system). These
bridges can also be utilized together with
VTT’s SULCA LCA tool.
Life cycle assessment is
a prerequisite for holistic
environmental evaluation
Life cycle assessment (LCA) is a compre-
hensive and ISO standardized method (ISO
14040:2006 [4], ISO 14044:2006 [5]) of evalu-
ating the environmental aspects and potential
environmental impacts of products. LCA can
also be applied in evaluating the impact of
technologies and processes. An LCA study
covers the whole life cycle of products, from
raw materials acquisition to end use, recy-
cling, or disposal. LCA provides information
to support decision-making in product and
technology development projects. LCA-based
information is applied in eco-labelling and the
production of environmental product declara-
tions. An LCA approach can also be applied
in the evaluation of eco-efficiency, material-
and energy-efficiency, and eco-design and life
cycle design.
Life cycle assessment has been devel-
oped in order to gain a better understanding of
the potential environmental impacts of prod-
ucts. As an example, LCA can be used for:
Catharina Hohenthal, Arto Kallio,
Marko Luukkainen, Tommi Karhela

32
• Identifying opportunities for improving
the environmental performance of prod-
ucts.
• Informing decision-makers in industry,
government or organizations.
• Selecting relevant indicators of the envi-
ronmental performance of products.
• Marketing products (e.g., making an
environmental claim or applying for an
eco-label or background information for
environmental product declaration) (ISO
14040:2006 [4]).
Each LCA study must be planned sepa-
rately and involves large-scale information
gathering. Thorough understanding of the
processes and products being assessed
is essential in order to enable evaluation
of the correctness of the data used in the
assessment as well as interpretation of the
assessment results. All products have some
impact on the environment, but improving
one part of the life cycle can also cause dete-
rioration in another. To be able to evaluate the
sustainability of a product or technology, sev-
eral indicators and extensive data collection
are required. An example of the life cycle of a
fiber product is presented in Figure 1, depict-
ing the complexity of such evaluation. Figure
2 shows the four stages of LCA; goal and
scope definition, life cycle inventory, impact
assessment and interpretation of results.
Benefitting from more than 20 years of
experience, VTT applies LCA in research
and customer projects in several sectors
of industry. VTT’s researchers participate
actively in the development of LCA method-
ology and tools in Finland and internationally.
We actively participate in ISO standardiza-
tion processes, including LCA (ISO 14040
series), carbon footprint (ISO 14067, 14069),
water footprint (ISO 14067), eco-efficiency
(ISO 14045) and social responsibility (ISO
26000).
Figure 1. Life cycle of a fibre product, showing inputs and outputs.

33
Software for life cycle assessment
The SULCA software allows the user to
perform life cycle inventory (LCI) and life
cycle impact assessments (LCIA) and to
present the calculation results in a clear,
easy-to-use way using unique reporting
features and configurable charts. Users
of SULCA enjoy connectivity with public
and commercial LCA databases, reducing
the effort required for data collection. With
this software calculation of carbon and water
footprints is easy and fast. The software –
SULCA4, sold internationally and currently in
use in more than 15 countries – is employed
by industry, universities, research institutes,
and others.
SULCA facilitates wider use of LCA by
lowering the level of effort required to con-
duct LCA studies. The tool is designed and
built in close cooperation with VTT’s LCA
experts using agile software development
methodologies to enable fast LCA model-
ling with less effort and lower cost. A key
aspect of the requirements engineering pro-
cess was the observation of LCA modellers
in their natural work environment during a
typical working day. As the end users tend
not to be software designers, they are less
able to explicitly describe their software
needs. Instead, this knowledge is obtained
by observing how users carry out LCA stud-
ies in practice. Observing the users enabled
often used features to be distinguished from
rarely used features and common usability
flaws and user mistakes to be identified. This
information is then used to prioritize software
requirements so that the most important fea-
tures of the tool are implemented first, in the
early stages of development. This gives the
users the opportunity to provide more feed-
back on the most commonly used features
and allows the software developers to build a
Figure 2. The four stages of LCA: goal and scope definition, life cycle inventory, impact
assessment and interpretation of results.

34
tool that better responds to the needs of LCA
professionals.
SULCA offers the following:
• Third-party database integrations
• Effective data management
• Support for KCL-ECO 4 models
• Simple and easy user interface
• Separate presentation for transport-
related elements
• Versatile configurability of flows allowing
closed loop systems
• Structural modelling
• Module classification
• Automatic unit conversions
• Mathematical formulas
• Impact assessment
Key features of SULCA include effective
data management, including sharing and re-
using data, and connectivity to public LCA
databases. Modellers are able to effectively
combine data from various sources, including
models from older software versions. Support
for mathematical formulas allows building of
configurable unit processes, reducing the
need to store numerous static configura-
tions. With global model parameters, users
are able to build multiple scenarios for a sin-
gle model for easy comparison. Transports
are represented with material flows between
processes to increase the clarity of the mod-
els. SULCA also includes a new intuitive
user interface with improved model valida-
tion, structural modelling and automatic unit
conversions. Advanced users can restruc-
ture the user interface by moving, detaching
and re-attaching UI components. The large
amounts of data connected with LCA model
simulations are represented in an easily inter-
pretable format. Modellers are able to quickly
find the information they are looking for and
make visualizations using the chart function.
With the help of SULCA’s module classifica-
tions, the user can easily identify which life
cycle stages cause the most environmental
burden.
LCA as part of an integrated
design and simulation environment
One of the key application areas of LCA is the
environmental impact assessment of emerg-
ing technologies. However, assessment is
often limited by a lack of robust data due to
the immaturity of these technologies. Process
simulation offers an interesting potential solu-
tion to this problem. Process simulation mass
and energy calculations can be combined
with LCA to support strategic decision mak-
ing regarding emerging technologies. This
approach has been suggested, for example,
by Liptow [6]. VTT has extensive experience
in both process simulation and LCA. In the
eEngineering spearhead programme we have
integrated our process simulation tools Apros
(www.apros.fi) and Balas (balas.vtt.fi) into a
common operating environment shared with
the SULCA LCA tool.
With the semantic data
transformation mechanisms of
Simantics we are able to transmit
data from process simulation
experiments to LCA experiments.
LCA is a widely-used technique for meas-
uring the environmental costs assignable to
a product or service. However, LCA takes
a high-level view and often assumes a fixed
supply chain structure and operation, with
sensitivity analyses often restricted to sce-
nario analysis of a limited number of possible
choices within this structure. Supply chain
design and practices can be a significant con-
tributor to overall environmental impacts. An
LCA approach typically considers the effect
of supply chain design and practices in ret-
rospect, with limited possibilities for ex ante
analysis of detailed process design options.
To overcome this problem, it has been sug-
gested, for example in [7] that LCA could be
combined with dynamic simulation. Using this
approach, environmental impact indicators

35
can be incorporated into a dynamic model
of the supply chain along with proﬁt and cus-
tomer satisfaction, so that the sustainability of
various design and operational decisions can
be assessed comprehensively. VTT is utilizing
system dynamics for modelling and simulation
of business processes. In the eEngineering
spearhead programme we have integrated
our System Dynamics tool (www.simantics.
org) and LCA tool SULCA into the same oper-
ating environment. This will enable combined
analyses in the future, as suggested above.
Towards integrated toolsets
Growing pressure on natural resources and
increased environmental awareness have
created growing demand for evaluating the
environmental aspects and potential environ-
mental burdens of products and services.
Against this background, the SULCA life cycle
assessment tool, used for conducting LCA
studies according to the ISO 14040:2006 and
ISO 14044:2006 standards ([4] and [5]), has
been developed in close collaboration with
VTT’s LCA experts to respond to their needs
and to enable LCA studies to be conducted
with less effort and cost.
Future work includes further usability
improvements based on increased user feed-
back once the tool is deployed for production
use. The new SULCA version is implemented
in Simantics – an open-source integration
platform for modelling and simulation tools.
Simantics enables connectivity with a grow-
ing set of other modelling and simulation tools
integrated into the same environment. Future
research includes exploiting the potential of
co-using LCA with process simulation, busi-
ness process simulation, and with system
dynamics and intelligent CAD environments.
For example, in process simulation an inte-
grated toolset enables environmental impacts
to be considered earlier in the process design.
Such advanced interdisciplinary research
would not be possible without the intensive
collaboration of various teams and knowledge
centres at VTT.
References
[1] Rockström, J., Steffen, W., Noonen, K. et
al. A safe operating space for humanity.
Nature 461, pp. 472–475.
[2] Ostad-Ahmad-Ghorabi, H., Collado-Ruiz,
D. & Wimmer, W. 2009. Towards Integrating
LCA into CAD. Proceedings of ICED 09,
the 17th International Conference on Engi-
neering Design, Vol. 7, Palo Alto, CA, USA,
24.–27.08.2009.
[3] Morbidoni, A., Favi, C., Mandorli, F. & Ger-
mani, M. 2012. Environmental evaluation
from cradle to grave with Cad-integrated
LCA tools. Acta Technica Corviniensis –
Bulletin of Engineering, Tome V 2012. ISSN
2067-3809.
[4] ISO 14040:2006. Environmental manage-
ment – Life cycle assessment – Principles
and framework. CEN international stand-
ards.
[5] ISO 14044:2006. Environmental man-
agement – Life cycle assessment
– Requirements and guidelines. CEN inter-
national standards.
[6] Liptow, C. & Tillman, A. 2011. Enhancing
the data basis for LCA through process
simulation: The case of lignocellulosic eth-
anol production in Sweden, SETAC Europe
21st Annual Meeting, 2011.
[7] Nwe, E., Adhitya, A., Halim, I. & Srinivasan,
R. 2010. Green Supply Chain Design and
Operation by Integrating LCA and Dynamic
Simulation. 20th European Symposium on
Computer Aided Process Engineering –
ESCAPE20. Pierucci, s. & Buzzi Ferraris, G.
(eds.). Elsevier B.V.
Karhela, T., Villberg, A. & Niemistö, H. 2012.
Open ontology-based integration platform for
modeling and simulation in engineering. Inter-
national Journal of Modeling, Simulation, and
Scientiﬁc Computing, Vol. 3, No 2, p. 36.
Finnveden, G., Hauschild, M., Ekvall, T.,
Guinée, J., Heijdungs, R., Hellweg, S.,
Koehler, A., Pennington, D., Suh, S. 2009.

36
Recent developments in Life Cycle Assess-
ment. Journal of Environmental Management
91, pp. 1–21.
Koukkari, H.& Nors, M. (eds). 2009. Life
Cycle Assessment of Products and Technolo-
gies. LCA Symposium. VTT Symposium 262.
VTT Technical Research Centre of Finland.
Espoo. 142 p.

37
Anyone dealing with modern digital equipment
is aware of the effects of inadequate software
quality. Software in smartphones and set-top
boxes has to be constantly updated – not to
introduce new features, but simply to manage
errors and issues as they emerge, to prevent
them from acting up or freezing on the user. But
the quality of the software of computer-based
systems that are critical, for example, to the
infrastructure of society, should be addressed
in much more serious ways. Almost every-
thing making up our infrastructure, from
road and rail traffic, supply chain logistics
and communication networks to power
grids and plants, is controlled, or at least
monitored, by software-based systems.
Although the design of industrial, criti-
cal software is based on completely different
practices to the software in consumer elec-
tronics devices, the goal of 100% error-free
software has nevertheless remained elusive.
But the situation is changing.
Model checking finds hidden
software errors
The process of analysing whether a system
design meets its requirements and fulfils its
intended purpose is called verification and
validation (V&V). For industrial instrumentation
and control (I&C) systems, V&V has tradition-
ally been based on testing and simulation – of
either an actual system or model replicas, i.e.,
running the target software and benchmarking
the behaviour and outcomes against sce-
narios or test cases. Test runs are a valuable
and necessary source of data, but testing and
Error-free software through formal methods
AUTHOR: Antti Pakonen
e-mail: antti.pakonen@vtt.fi
simulation alone cannot be relied on to prove
that a system is 100% error-free. Test spaces
easily grow to immense proportions so that, in
practice, not all possible test cases or scenar-
ios can be taken into consideration. Specific,
advanced test automation tools are available,
but conclusive analyses are impossible due to
the sheer number of possible situations that a
system can, in theory, and with the necessary
critical approach, be shown to incur.
Model checking is a computer-assisted
formal method that can prove conclusively
whether a (hardware or software) design
model acts according to its stated require-
ments in all situations. Both the system design
and the requirements are presented in a for-
mat understood by a model checking tool,
called a model checker, which will then thor-
oughly analyse whether a model execution
that is contrary to the requirements is possi-
ble. As a general principle, instead of looking
at what happens in a given situation with given
inputs, the idea is to define an undesired situ-
ation and see if it is possible to end up there.
Instead of excessively computing all com-
binations, a systematic search is carried out
using a graph-like model, looking only at the
combinations that are relevant to each stated
requirement.
Since the 90s, formal model checking has
been the key verification method in micropro-
cessor manufacturing, and has recently found
its way into ever more versatile domains. At
VTT, we have been applying model checking in
the V&V of critical I&C software, i.e., the logic
that controls and monitors industrial processes.
Janne Valkonen

38
Developing practical tools for
industry
Versatile and mature model checking tools
are available, but most are either too generic
and abstract or aimed at a specific domain
and, therefore, not suitable for the analysis
of I&C software. So-called function blocks
are one of the most common programming
languages used to implement I&C sys-
tems. Instead of writing code, applications
are constructed by selecting predefined
standard blocks (e.g., AND, OR, delay, or
a PID controller) and connecting (‘wiring’)
them together in a graphical diagram to
obtain the desired functionality or operation
logic. Each block reads its inputs, updates
its internal variables, and sets its outputs
according to its internal logic. Block wiring
then defines the data flow and the block pro-
cessing order. Function block diagrams are
favoured, since (among other benefits) they
present clear input-output mapping, and it is
relatively easy to understand and follow the
processing flow.
Accordingly, VTT has been
developing tools for model checking
of function block based software.
Our work has been based on
Simantics, an open source platform
for modelling and simulation.
We have been specifying a
Simantics plugin for the open
source model checker NuSMV,
and are now able to construct the
NuSMV model by wiring blocks
together in a 2D graphical view.
The expected benefits of such a graphi-
cal, dedicated toolset are clear. In the future,
the model translation capabilities of Siman-
tics are also expected to enable automatic
model conversion from, for example, an
existing Apros model of process control
software.
Practical experience in evaluating
industrial control systems
The majority of VTT’s practical experience
in applying model checking is in the nuclear
industry (see Case: Nuclear on the next
page) where very strict safety analysis pro-
cedures are an essential requirement. The
approach is, however, also more generally
applicable, and we have conducted small-
scale pilots in diverse I&C applications in
which different programming languages and
environments from different vendors have
been used. Successful pilots have been per-
formed, for example, in factory, power plant,
electrical and machine automation projects.
The model checking approach is most
suitable for the analysis of relatively straight-
forward logic – the kind of logic that should
be favoured for safety-critical applications
– as the computational power of model
checking is based on the use of fairly sim-
ple modelling languages. Conversely, more
sophisticated and algorithmically complex
control applications, such as those needed
to run a modern paper machine, for exam-
ple, cannot be effectively analysed, in
which case, for example, simulation-based
verification and validation are more suit-
able. Nevertheless, just because a system
is straightforward does not mean that its
analysis is simple: a binary circuit with 100
inputs and no internal memory will have 1030
different input combinations, and adding
memory to the application only further com-
plicates the analysis.
Solving the theoretical challenges
VTT has been working with Aalto University
to solve some of the theoretical challenges
related to model checking and, in particular,
the evaluation of I&C software. The greatest
challenge is the computational effort required
due to the state explosion problem: as the
number of possible model states grows expo-
nentially with respect to the size of the model,
the analysis task can become too complex
for existing methods and computers. Model

39
CASE: NUCLEAR
With new-builds and modernizations, old analogue nuclear power plant technology
is being steadily replaced by digital instrumentation and control (I&C) systems. Software-
based control systems can offer higher reliability, better plant performance and new
diagnostic capabilities. Nevertheless, the inherent complexity of digital I&C has justiﬁably
raised questions regarding the correctness of both hardware and software design. The
industry and regulators thus face important challenges in assuring that new systems meet
their requirements.
At VTT, research on model checking began in 2007 under the Finnish Research Pro-
gramme on Nuclear Power Plant Safety (SAFIR2010). Successful industrial pilot cases
quickly proved the value of the approach. Finnish nuclear power companies and authori-
ties have shown continued interest in formal methods, and research continues under the
SAFIR2014 programme.
In addition to active research, the approach has been put to practical use. VTT has
been consulting the Finnish Radiation and Nuclear Safety Authority (STUK) (since 2008)
and the power company Fortum on evaluating nuclear I&C systems using model checking.

40
checking is a computationally very powerful
method, but it, too, has its limits.
Specific topics for past and present
research include:
• Modular approach to analysis of very
large models: A technique has been
developed to analyse systems that would
otherwise result in models with a too
large state space, based on the modular
structure of the model. The model is first
approximated by greatly abstracting the
logic of some of the modules. An algo-
rithm has been developed that iteratively
searches for a composition of modules
that at the same time is computation-
ally manageable, and covers enough
modules to prove the properties of the
original model. The feasibility of the anal-
ysis results for the abstracted model is
then reviewed in the context of the full
model.
• Analysis of asynchronous models: For
many model checkers, one of the nec-
essary simplifications needed in order
to make the analysis efficient is that the
analysed system model has a unified,
discrete time cycle. However, many real-
world systems are physically distributed
to several different processors that each
behave according to their internal clock.
Model checkers that can also handle
asynchronous behaviour do exist, but in
these cases the size of model that can be
effectively analysed is clearly smaller. A
new model checking tool is being devel-
oped that will combine the strengths of
different types of model checkers.
• Modelling system faults: If the I&C soft-
ware has a mechanism for dealing with
faulty input data, the mechanism is taken
into account in the model. However, if we
wish to introduce failure modes of the
underlying hardware architecture (‘What
if one of the processors the software is
running on or the communication network
fails?’), extra work is needed in defining
suitable failure mechanisms. We are cur-
rently working on structured approaches
for doing so.
• Reliability through tool diversity: When
a model checker discovers no error,
there is always some question whether
the design actually is error-free. Errors in
the modelling process are usually found,
and most often result in a ‘false negative’
result. One way of increasing confidence
in the results is to use several model
checkers that do not share source code
in their implementation. We are currently
constructing a tool portfolio that will not
only enable the evaluation of more versa-
tile applications than before, but also add
to the reliability of analysis results.
Summary and need for further
work
Through model checking, we have been
able to find hidden design errors in software
systems that have already undergone verifi-
cation and validation through more traditional
means. Others have reported similar results
in diverse application areas, such as aviation.
The method is not, however, a one-size-fits-
all solution, since it is only effective for the
evaluation of fairly straightforward software
applications (which safety-critical industrial
control systems often are). Also, expert knowl-
edge is always needed when applying formal
methods.
VTT is currently working on the theoreti-
cal challenges as well as more practical issues
related to the application of model checking
in industrial contexts. The Simantics plat-
form enables us to bring model checking
to the mainstream, as the dedicated, user-
friendly tool makes it possible to implement
model checking with less knowledge of the
underlying theory. Our current tool devel-
opment approach is tied to function block
diagrams as the programming language of
I&C software, but other viewpoints are also
needed, as, for example, the C language is
often used in the industry, and model check-
ers for verifying C code are also available. On

41
the theoretical side, new methods are needed
to handle the specification of system require-
ments. Our ongoing research is nevertheless
motivated by successful applications both in
research pilot cases and in practical customer
projects.
To date, practical application of VTT’s
model checking has been mainly within
the context of nuclear power plants, as the
nuclear industry is subject to rigorous legis-
lative requirements regarding safety analysis.
However, safety is not the only criterion driv-
ing strict V&V – cost is also an important
factor. While the expertise needed for model
checking does not come free of charge, the
expenses caused by the downtime of indus-
trial plants or infrastructure systems due to
design faults can be immense.
Lahtinen, J., Valkonen, J., Björkman, K.,
Frits, J. & Niemelä, I. 2012. Model checking
of safety-critical software in the nuclear engi-
neering domain. Reliability Engineering and
System Safety, Vol. 105, pp. 104–113.
Pakonen, A., Mätäsniemi, T. & Valkonen,
J. 2012. Model Checking Reveals Hidden
Errors in Safety-Critical I&C Software. 8th
International Topical Meeting on Nuclear
Plant Instrumentation, Control, and Human-
Machine Interface Technologies (NPIC & HMIT
2012). San Diego, California, USA, 22–26 July
2012. American Nuclear Society. Pp. 1823–
1834. ISBN 978-0-9448-093-0.
Lahtinen, J., Launiainen, T., Heljanko, K.
& Ropponen, J. 2012. Model Checking
Methodology for Large Systems, Faults and
Asynchronous Behaviour. Espoo: VTT. 84 p.
(VTT Technology 12.) ISBN 978-951-38-7625-
8. http://www.vtt.fi/inf/pdf/technology/2012/
T12.pdf.
Lahtinen, J., Valkonen, J., Björkman, K.,
Frits, J. & Niemelä, I. 2010. Model check-
ing methodology for supporting safety critical
software development and verification. Euro-
pean Safety and Reliability Conference
(ESREL2010). Rhodes, Greece, 5–9 Sept
2010. Ale, B.J.M., Papazoglou, I.A., Zio, E.
(Eds). Reliability, Risk and Safety – Back to
the Future. European Safety and Reliability
Association (ESRA). Pp. 2056–2063. ISBN
978-0-415-60427-7.
Valkonen, J., Björkman, K., Frits, J. & Nie-
melä, I. 2010. Model checking methodology
for verification of safety logics, The 6th Inter-
national Conference on Safety of Industrial
Automated Systems (SIAS 2010), Tampere,
Finland, 14–15 June 2010. http://www.vtt.fi/
inf/julkaisut/muut/2010/SIAS_final.pdf.

42
Authors
Susanna Aromaa Marko Antila
Research Scientist Senior Scientist
Eero Kokkonen Boris Krassi Simo-Pekka Leino
Senior Scientist Senior Scientist Senior Scientist
Hannu Nykänen Kaj Helin Sauli Kiviranta
Principal Scientist Principal Scientist Research Scientist

43
Designing user experience
for the machine cabin of the future

45
Designing user experience for the machine cabin of the future
In addition to safety, operator comfort has
become an essential driver of product design
and development in the mobile machine
industry. Typically, the machine cabin designer
focuses on one or, at most, a few proper-
ties at a time, such as physical ergonomics
or machine feasibility during task execution.
However, user experience (UX) and comfort
are a combination of several different fea-
tures. A holistic UX and comfort evaluation
includes aspects such as psychoacous-
Designing user experience for the machine
cabin of the future
AUTHOR: Susanna Aromaa
e-mail: susanna.aromaa@vtt.ﬁ
tics, thermal comfort, vibrations within the
cabin, musculoskeletal discomfort, and
visibility from the cabin. For many of these
factors, UX vary greatly between individuals,
tasks and working environments. New tools
and methods are therefore needed to iden-
tify and optimize these factors to meet user
requirements in design and to evaluate their
feasibility in practice.
But how to evaluate human–machine
interaction parameters and their combined
Figure 1. The multidisciplinary design environment enables designers to take key factors
affecting user experience and comfort into early consideration in the design process.
Marko Antila, Eero Kokkonen,
Boris Krassi, Simo-Pekka Leino,
Hannu Nykänen, Kaj Helin, Sauli Kiviranta
Integration
Air conditioning
modelling
User experience
evaluation
Virtual
environment
Acoustic
modelling
Design Environmet for User Experience
modelling evaluationenvironmentmodelling

46
effect on the UX in the early design phase, when
physical prototypes are too costly to construct
or not available? This can be achieved by
using Virtual Environments (VEs), which com-
bine design models and operation simulations
to enhance the natural feel of a simulated work
task, when evaluating a design. With VEs
that include a good visualization system,
realistic motion platform, realistic acous-
tics, and air conditioning modelling, it is
possible to assess the combined effect of
these parameters, thus benefitting the user
through improved design with respect to
comfort, ergonomics, usability and safety.
In addition, machine manufacturing com-
panies also benefit through reduced design
costs, faster time-to-market, and better prod-
uct quality, as requirements can be evaluated
early and modifications can be made fast.
Virtual design environment enables
holistic consideration of the user
The main outcome of this sub-project of the
eEngineering programme was the develop-
ment of a virtual, simulator-based design
environment for the multidisciplinary UX
design of machine cabins. The project had
two main objectives: (1) Integration of vari-
ous areas or systems, such as, acoustic
modelling, air condition modelling, and UX
evaluation into one design environment (such
as VEs), see Figure 1. (2) Development of
UX and comfort evaluation methods (psy-
choacoustics, thermal comfort, whole-body
vibration, musculoskeletal discomfort and
operator’s field of view) to enable more holis-
tic UX evaluation.
Holistic user consideration
changes the entire product design
and development process
Impacts on product processes were also
studied during the project, with a focus on
how use of the novel virtual design environ-
ment changes current product design and
development processes, and how to bring the
environment into normal practice within indus-
try. Product or system design life cycles – from
Figure 2. Impacts of the multidisciplinary design environment on virtual and physical
product life. The left side illustrates the phases of the design process. The use of VE
smooths the iterations and phases and eases communication across phases. The right
side shows the effect of VE on time-to-market due to earlier and more effective require-
ments validation and verification and thus fewer engineering changes and interruptions.

47
product specification and business targets
to detailed verified technical solutions – are
often described using the systems engineer-
ing V-model. Adapted V-model (see Figure
2) illustrates how integrated product devel-
opment simulators bring improvements and
significant changes to the product design and
development process.
Figure 2 shows how the use
of virtual simulators shifts the
actual, physical system towards
earlier commercial product launch
compared to traditional engineering.
It enables earlier and better
decision making based on earlier
evaluation and validation of user
and other stakeholder requirements,
and verification of combined
multidisciplinary design solutions
with less engineering changes
during product development and,
therefore, faster time-to-market.
Experiences from our partners show
that impacts are actual.
Capturing the user experience in
the virtual environment
UX and comfort are essential factors in high
quality cabin design. UX can be defined as
a person’s perceptions and responses
resulting from the use of a product, sys-
tem or service. Additionally, comfort is
defined as a subjective, personal experi-
ence, affected by various factors (e.g. touch,
sight, hearing, taste and smell) and reaction
to the environment [1].
VEs help designers to get understand-
ing of the UX and to design systems that
take into account human needs while ensur-
ing that the cognitive and physical potential
of the user are utilized with respect to the
overall goals of the system. In addition, it
can ensure, during the design, the recogni-
tion of different users involved to product life
cycle such as assembly, maintenance and
operation. Another benefit of VEs and virtual
reality (VR) technologies is that they enable
fast comparison of radically different design
Figure 3. The creation of virtual environments through interaction between different
sub-environments. The left side shows the relationships between the parts of the virtual
environment (ref. [2, p. 6]), the right side shows the basic technology enablers for a virtual
environment (also shown in Figure 4).
virtual environments through interaction between different
Visual
Environment
Auditory
Environment
Virtual
Environment
Haptic/
Kinaesthetic
environment
1. Visualization system with active stereographic
rendering in three screens power wall setup or
with Head Mounted Display (HMD)
2. Marker-based optical motion capture system in
order to caprure user motions
3. User interface systems e.g.different
gaming or real controllers
4. Surround audio system
with headphones
5. Motion platform to
replicate the
machine’s motion

48
solutions without the need for physical pro-
totypes.
The main quality criteria for VEs have tra-
ditionally been the quality of 3D graphics and
the generation speed and smoothness of visu-
alization. High quality visuals are undoubtedly
the single most important factor when aiming
to create an effective and functional VEs. How-
ever, 3D graphics alone are not sufficient to
produce a realistic illusion of an environment.
In addition to vision, manipulation of the other
senses is also required to create a convinc-
ing illusion of reality, i.e., VR. The key senses
include hearing, haptic (touch) and kinaes-
thetic (body position and movement) senses,
and to a lesser degree, taste and smell.
Figure 3 [ref. 2, p. 6], depicts how a VE
consists of several simultaneously interacting
Figure 4. The virtual environment developed at VTT consists of five subsystems: (1) Main
visualization system with active stereographic rendering in three walls setup plus sec-
ondary visualization system with head-mounted display (HMD); (2) Marker-based optical
motion capture system to capture user motions; (3) User interface (UI) system combining
different gaming controllers and basic keyboard interaction; (4) Ambient audio system
with headphones; and (5) Motion platform to replicate machine motion.
sub-environments. The visual, auditory, and
haptic environments together form the VE.
If any of these is missing, the VE is consid-
erably less functional and realistic. The VE
created at VTT and its subsystems are illus-
trated in Figure 4.
Development of the UX and comfort
evaluation methods was started by selecting
key cabin design parameters, such as: psy-
choacoustics, thermal comfort, whole-body
vibration, musculoskeletal discomfort and
operator’s field of view.
Psychoacoustic Experience Evaluation
and Enhancement (PEEE) is a method for
evaluating how human beings experience the
acoustic environment and for improving key
factors of the acoustic experience. Real or
modelled sound events in a real or modelled
(2)
(1)
(3)
(5)
(4)

49
acoustic environment are captured or gen-
erated in binaural form. Binaural signals are
then used in listening tests or for extraction
of individual psychoacoustic metrics, such as
loudness and sharpness.
UX of thermal comfort is evaluated by
applying Fanger’s thermal comfort model [3,
4]. Additionally, window fogging and dust dis-
persion can be simulated. Air flows and other
cabin air conditioning related phenomena can
then be visualized in VEs enabling users to
visually experience air flow streams, thermal
comfort, window fogging and dust dispersion.
In order to measure the whole-body vibra-
tion dose experienced by the user, data based
on standard acceleration is gathered from the
motion platform or by collecting acceleration
sensor data from the seat. The vibration dose
is calculated based on standard ISO 2631-1
[5].
Subjective musculoskeletal symptoms of
the user are collected via a computer-aided
tool that enables the user to choose a body
part from a body map and then indicate their
severity of discomfort. Data on experienced
musculoskeletal discomfort and degree of
discomfort is gathered before and after per-
forming the task in the VEs.
To increase the use of real operators in
field of view (FOV) evaluations, a new method
for calculating FOV in VEs was developed.
The FOV analysis method is based on task-
related visibility and occlusion evaluation
of target objects in the operator’s FOV. The
method calculates (1) the percentage of vis-
ible target object pixels from all pixels in the
operator’s FOV, and (2) the percentage of
occluded (by the cabin structure) pixels from
the visible target object pixels in the opera-
tor’s FOV. The results facilitate comparison of
the impacts of alternative design solutions on
visibility.
Auralization: the virtual hearing
experience
Effective audio is key to building a convinc-
ing VR experience. In VEs, sound events and
acoustics are simulated to sufficient accu-
Figure 5. Modular sound generation for virtual reality. Position tracking follows the posi-
tion and movements of objects (including the observer) in the VE. A dynamics solver
creates the dynamic parameters for the model. This information is forwarded to the VR
software, which translates them into an appropriate format for the sound creation block.
Sound from the sound creation block is localized by the sound localization block, and
then sent either to the headphones or loudspeakers.
POSITION
TRACKING
DYNAMICS
SOLVER
VIRTUAL
REALITY
SOFTWARE
Audio Sub-system
Sound
creation
Sound
localization
5.1
Loud-
speakers
Head-
phones

50
racy to create an immersive experience while
not placing excessive demands on system
resources. The most important requirement
is real-time and concurrent operation of the
audio simulator and visualization to preserve
the illusion of immersion. Visualization events
must be tightly synchronized with the aural-
ized environment.
In general, sound events and acoustics
form an audio sub-system itself, as illustrated
in Figure 5. Sound events are generated in
the Sound Creation block and are positioned
accordingly in the Sound Localization block.
A 5.1 loudspeaker setup comprising 3 loud-
speakers in the front (left, right, and middle),
2 rear loudspeakers and a subwoofer, was
used as the main sound source. Panned
sound within these loudspeakers can also be
converted to headphone use. Sound panning
involves projecting sound in a certain direction
by changing the level of sound in each indi-
vidual loudspeaker. The audio subsystem gets
its parameters in real time from the VR soft-
ware, which generates the parameters based
on position tracking and information from a
dynamics solver.
Sounds can be created in the VEs in
various ways. In many industries the common
practice is simply to pre-record or sample
noise signals and then play them back in the
VEs. At the other end of the spectrum, audi-
ble models are being developed to generate
sound and noise based on parameters such as
tonal component levels, frequencies, relative
phases, broadband noise frequency content,
and level of amplitude modulation. Such audio
models offer the potential to create truly vir-
tual audio experiences and are therefore a key
focus area of current research.
The Audible Model Platform
(AMP) developed by VTT is a
parameter-based sound and noise
generation platform. The AMP is
not directly based on the structural
or mechanical physics of noise
generating machinery, but rather
the noise proﬁle parameters are
functions of angular velocity (rpm),
load, listening position and other
relevant factors.
Figure 6. Audible Model Platform (AMP) tailored for the cabin noise model. The sources
of noise are engine orders, hydraulics orders, engine broadband noise and ventilation
noise. All of these are affected by various parameters coming from external VE source.
Enable Sub-system
3D
Head-
phones
Sound
localization
Engine
orders
Aux
periodic
sources
(hydraulics
turbo ect.)
Engine
broadband
Order
scaling
Order
scaling
VENTILATION
NOISE
RPM/LOAD/
OTHER
PARAMETERS
SOURCE
5.1
Loud-
speakers

51
The AMP is used here for cabin noise
modelling, but it has also been customized
for various fixed engine applications and even
outdoor noise applications, most recently for
noise modelling of wind turbines. AMP noise
profile parameters include engine and other
rotating mechanical system acoustical order
structure, levels and phases, as well as, similar
parameters for auxiliary mechanical systems,
such as cooling, turbo and hydraulics. Fur-
thermore, broadband noise is parameterized
whenever applicable, and otherwise modelled
based on a steady-state empirical model. Due
to the parameter-based design, changes in
parameters also cause real-time changes
in the rendered audio.
The noise profile parameters can be
extracted from the measurements or noise
recordings. For this extraction, automatic and
semi-automatic tools have been generated.
Parameters from the numerical models can
also be used if they can be generated. The
AMP also visualizes engine orders and gen-
erated noise, and the visualization parameters
can also be sent back to the VE. Each order
contribution is visualized in real time, as well
as the overall noise spectrum. Individual nar-
rowband or broadband components can be
turned on and off to evaluate their noise con-
tribution.
An example AMP for cabin noise genera-
tion is presented in Figure 6. Here, the main
sources of noise are engine orders, hydraulics
orders, engine broadband noise and ventila-
tion noise. All of these are affected by various
parameters coming from external VE source,
such as dynamics solver. The noise level and
sound quality parameters are also calculated
and are available in the model or as outputs
to VR software.
Air conditioning in a virtual cabin –
more than just temperature control
Detailed information on air flows and other
cabin air conditioning related phenomena
can be obtained using computational fluid
dynamics (CFD). For example, air flow pat-
terns, thermal comfort, window fogging
and dust dispersion inside the cabin can
be determined computationally without
a physical model of the cabin, as dem-
onstrated in Figure 7. The cabin geometry
and details such as the location of supply
air inlets can be varied and the effects of the
Figure 7. Computational fluid dynamics (CFD) simulation can be executed at the very
beginning of the product development cycle to provide detailed information on air con-
ditioning properties from the user’s perspective even if no physical models of the cabin
exist. The resulting air flows, thermal comfort, window fogging and hazardous silica dust
dispersion to breathing zone of the operator can all be visualized in 3D.

52
changes can be estimated. All of this can be
carried out at the very beginning of product
development, providing much more detailed
knowledge of the system compared to what
physical models alone would provide.
The UX related to air conditioning is
highly dependent on cabin airflow patterns,
which in turn are dependent on, for example,
the geometry of the cabin. Detailed data on
the steady-state flow field (i.e. velocity, pres-
sure, temperature, etc., values in each spatial
location) are obtained from the CFD calcula-
tion. In the standard procedure of CFD, the
domain of interest, here a work machine
cabin, is divided into a grid of small com-
putation or control volumes over which the
Navier–Stokes equations are solved numeri-
cally, resulting in 3D data for the flow field. The
thermal comfort of the operator can be evalu-
ated by applying this CFD data to Fanger’s
thermal comfort model. Thermal comfort
is determined based on six parameters: air
temperature, air velocity, mean radiant tem-
perature, relative humidity, metabolic activity
level and clothing of the human. The comfort
indices are widely used and have reached
the status of normative reference (ISO 7730).
The classical comfort indices are suitable
Figure 8. The result of an air conditioning simulation comprising the evaluation of thermal
comfort, window fogging and dust dispersion from floor by air flows. The experienced
thermal environment (average over a group of people) is represented by PMV (predicted
mean vote) ranging from cold (-3) via neutral (0) to hot (+3). Humidity originating from the
operator, wet floor and supply air inlets, dispersed and condensated on the windows,
is visualized by the variable filmMass (kg/m²/s). The dust concentration dispersion from
the floor is visualized by constant concentration contours. Instead of analysing absolute
concentration values, the resulting concentration values are normed to the uniformly
dispersed concentration; for example, a scaled concentration of dust scaled = 1 corre-
sponds to the dust concentration if it would be uniformly mixed throughout the whole air
volume of the cabin.

53
for representing conditions in an enclosed
space provided that fairly uniform conditions
hold. However, the airflow in a cabin envi-
ronment does not satisfy the assumptions
of uniform conditions. Contrary to Fanger’s
original method, heat exchange is modelled
by CFD methods and the actual flow proper-
ties provided by the CFD calculation are used
to determine the comfort indices. In addi-
tion, fogging of the cabin windows can be
modelled by solving the dispersion of water
vapour inside the cabin and by solving the
condensation onto the window surface. Fur-
thermore, the transport of hazardous silica
dust into the breathing zone of the operator
can be modelled by solving the dispersion of
dust from the floor due to air flows. In some
cases, such as an underground loader, the
cabin floor can become covered with sand
containing hazardous quartz (SiO2) dust
particles. An example of air conditioning sim-
ulation is shown in Figure 8.
Empowering manual work with
augmented reality
According to Eurostat [6], in 2012 15.7 mil-
lion people1
were involved in high-knowledge
manual work in Europe, mainly as plant and
machine assemblers and operators. Numer-
ous industry sectors depend on the knowledge
and skills – such as satellite assembly, nuclear
reactor maintenance, operation of complex
machinery, design and manufacturing of highly
customized products – of their manual work-
ers. In these sectors, manual work constitutes
the core operations and cannot be off-shored
or easily automated.
The EU-funded ManuVAR project
coordinated by VTT [7, 8] developed a new
system with the potential to significantly
improve productivity and working environ-
ments across Europe. The project combined
product life cycle management, ergonomics,
and virtual and augmented reality (VR and AR)
technology.
The following four main results were
achieved by the project.
1. Most prominent problem areas faced
by European industries in the context
of high knowledge high value manual
work:
* Hindered communication across vari-
ous actors throughout the life cycle.
* Poor interfaces with complex CAD
and information systems.
* Inflexible design processes – feedback
from later life cycle stages is difficult to
utilize in design improvement.
* Inefficient knowledge management –
substantial employee know-how not
utilized in system design and improve-
ments.
* Low productivity of manual work due
to poor overall system design.
* Lack of acceptance of supporting
technologies, especially virtual and
augmented reality technologies, in
industrial contexts.
* Physical and cognitive stress of man-
ual workers, could be minimized by a
better system design.
2. Five industrial cluster cases including
performance criteria for the evaluation of
laboratory trials and factory-floor dem-
onstrations, business analysis, economic
impact forecast, training and technology
transfer plans:
* Cluster 1: Support of Spacecraft
Assembly. Develop and validate criti-
cal procedures in VR that can be used
to support integration assembly activi-
ties through AR instructions.
* Cluster 2: Manufacturing design for
SMEs. Support assembly line workers
by means of an automatic work load
evaluation tool and reduce learning
time by means of an operator naviga-
tion tool.
* Cluster 3: Remote support in train
maintenance. Support the main-
tenance of complex systems by
exploiting the benefits of AR technol-
ogy and reinforcing communication
between actors involved.
1
Total obtained for EU27 countries from the referred table “Employment by sex, age, professional status
and accupation” as of 04-07-2013, by selecting industry category (ISCO) “Plant and machine operators,
and assemblers”, data for 2012.

54
* Cluster 4: Training for industrial plant
maintenance. Training for metallo-
graphic replica activities using VR with
visual, audio and haptic interaction.
* Cluster 5: Design and maintenance
of heavy machinery. Assembly and
maintenance design reviews and
instructions.
3. System architecture characterized by the
following features [9]:
* Bi-directional communication
throughout the system life cycle
(e.g. worker feedback to designers,
designer recommendations to work-
ers) is accomplished by means of a
‘virtual model’. The virtual model plays
the role of communication media-
tor – a single systemic access point
to a variety of system data, informa-
tion and models for all users in the life
cycle – accessed as an integral sys-
tem by ‘virtual experiments’;
* Adaptive VR/AR user interfaces to the
complex virtual model, suitable for all
actors in the life cycle: from workers
to engineers to managers. The VR/AR
interfaces are implemented by com-
ponent reconﬁguration with low-delay
middleware (haptics, tracking, VR/AR
Contextual instruction delivery: AR,
tracking; remote and local versions
Figure 9. ManuVAR application tools. Source: ManuVAR consortium.
Ergonomics evaluation: automatic physical and cog-
nitive load analysis, full body motion capture
Task analysis and procedure validation: hierar-
chical task analysis, VR with haptics
Motor skill training: VR with haptics, precision
teaching theory

55
visualization, application logic, con-
nection to PLM systems);
* Four groups of ergonomics meth-
ods covering the principal ways of
improving manual work from the
system-cybernetics perspective:
workplace design, ergonomics evalu-
ation, instruction delivery, and training;
* Knowledge management concept
based on Nonaka’s organizational
knowledge creation theory, with
each modality of knowledge creation
supported: externalization and inter-
nalization (adaptive and natural user
interfaces with VR/AR), socialization
(bi-directional communication and
the virtual model), and combination
(linking in the virtual model and con-
nection to PLM systems).
4. Four reconfigurable application tools,
which can be combined together via the
virtual model to solve a given industrial
case, were designed, implemented and
evaluated in the laboratory and in the
company environment, Figure 9.
The ManuVAR system was demon-
strated in all five project clusters. Around 110
workers, engineers, managers and custom-
ers from 23 external companies were also
included in the project. The feedback was
constructive and indicated considerable
interest from the industry:
• ‘I think training could be performed in
less time, reducing the “in-class” training
of trainers, and so be much more effi-
cient with ManuVAR’.
• ‘Operators will be more open to record-
ing and analysing their postures and
movements. Giving immediate results
and feedback will probably encour-
age them to change their postures and
movements’.
• ‘Connection to the company PDM would
make it possible to use simulations and
modify data models using an innovative
recursive process rather than the normal
waterfall approach’.
Towards user experience design
The multidisciplinary UX based approach to
cabin design implemented in the eEngineering
programme appears, based on the first ver-
sion integration alone, to be highly promising.
The results also appear generalizable to the
transportation and manufacturing industries.
In the future, finding ways for reliable evalua-
tion of comfort continues to be challenging as
well as overcoming technical constraints of vir-
tual reality systems that may limit the realistic
user experience. Nevertheless, the presented
approach to cabin design is very beneficial
due to its holistic approach to these complex
socio-technical systems.
References
[1] De Looze, M.P., Kuijt-Evers, L.F.M. & Van
Dieën, J.H. 2003. Sitting comfort and
discomfort and the relationships with
objective measures. Ergonomics, Vol. 46,
pp. 985–997.
[2] Kalawsky, R. S. 1993. The Science of
Virtual Reality and Virtual Environments.
Addison-Wesley.
[3] Fanger, P. 1967. Calculation of thermal
comfort: introduction of a basic comfort
equation, ASHRAE Trans., 73, III.4.1–
III.4.20.
[4] Fanger, P. 1970. Thermal comfort, Copen-
hagen, Danish Technical Press.
[5] ISO 2631-1. 1997. Mechanical Vibration
and Shock – Evaluation of Human Expo-
sure to Whole–Body Vibration. Part 1:
General Requirements. Geneva.
[6] Eurostat: http://appsso.eurostat.
ec.europa.eu/nui/show.do?dataset=lfsa_
egais&lang=en
[7] ManuVAR web site: www.manuvar.eu
(accessed August 6, 2013)
[8] Krassi, B. 2012. Manual work support
throughout the system life cycle by
exploiting Virtual and Augmented Reality
(ManuVAR). In ‘Production matters: VTT in

56
global trends’. Häkkinen, K. (ed.). Espoo:
VTT. Pp. 84–88, ISBN 978-951-38-7860-3.
www.vtt.fi/inf/pdf/researchhighlights/2012/
R3.pdf
[9] Krassi, B., Kiviranta, S., Liston, P., Leino,
S.-P., Strauchmann, M., Reyes-Lecuona,
A., Viitaniemi, J., Sääski, J., Aromaa, S.,
Helin, K. 2010. ManuVAR PLM model,
methodology, architecture, and tools for
manual work support throughout system
lifecycle. Proceedings of the 3rd Inter-
national Conference on Applied Human
Factors and Ergonomics (AHFE2010),
Miami, Florida, USA.ISBN-13: 978-0-
9796435-4-5.
Aromaa, S., Leino, S.-P., Viitaniemi, J.,
Jokinen, L. & Kiviranta, S. 2012. Benefits
of the use of Virtual Environments in product
design review meeting. International Design
Conference, Dubrovnik, Croatia, May 2012,
21–24. 8 p.
Aromaa, S., Leino, S.-P., Kiviranta, S.,
Krassi, B. & Viitaniemi, J. 2012. Human-
machine system design: the integrated use
of human factors, virtual environments and
product lifecycle management. Tijdschrift voor
Ergonomie, Vol. 37, No. 3, pp. 11–16.
http://www.vtt.fi/inf/julkaisut/muut/2012/
Human-machine_system_design.pdf
Viitaniemi, J., Aromaa, S., Leino, S.-P.,
Kiviranta, S. & Helin, K. 2010. Integration of
User-Centred Design and Product Develop-
ment Process within a Virtual Environment.
Practical case KVALIVE. Espoo, VTT. 39 p.
VTT Working Papers; 147. ISBN 978-951-38-
7489-6.
http://www.vtt.fi/inf/pdf/workingpapers/2010/
W147.pdf

58
Impacts of eEngineering 2009—2012
Inputs • The volume of the programme EUR 31 million, approximately 250 person
years.
o The annual budget varied between EUR 7-9 million.
• The programme consisted of ca. 130 projects:
o 43 contract research projects, with total revenue of 6,1 M€.
o 63 public and jointly funded projects, totalling 18,4 M€ (13,6 M€ external
funding and 4,8 M€ VTT’s own funding).
o 15 of the joint projects were funded by EU (with EU revenue 3,7 M€).
o 11 of the joint projects were carried out in the context with SHOK’s
(Strategic Centres for Science, Technology and Innovation) and 9 pro-
jects in the context with SAFIR (Finnish public research programme on
nuclear power plant safety), while others were funded directly by Tekes
or Academy of Finland.
o 30 projects self-funded by VTT with total revenue of 6,7 M€.
o Included in the above figures, there were three multi-million project clus-
ters of several industrial and university partners, each contributing with
their own resources.
Out-
puts
• Several successful technology transfer actions, for example,
o 87 licensing agreements involving simultaneous process development
with the client
o Contracts with 5 big industrial customers, reflecting the commercial rel-
evance of programme themes
o Formation of 2 alliances of private and public partners
• 5 invention disclosures
• 1 start-up company (Semantum)
• Altogether ca. 40 technical and scientific (10) publications, including peer
reviewed scientific journals, international conference papers, and VTT pub-
lications.
• Development and release of the SIMANTICS platform, an ontology based
integration environment for modelling and simulation, that enables the
linking and co-use of models of different levels of details through different
viewpoints to the models and transformations between engineering (CAD)
information and simulation models.
• Simantics Constraint Language (SCL) for transformations between engi-
neering and simulation components’ data models within SIMANTICS.
• Development and release of several other simulation and engineering tools
• 2011 Automation award of the Finnish Automation Society to VTT’s Siman-
tics team
• 2010 VTT Award to Research Professor Tommi Karhela for outstanding
work on the Simantics platform

59
High-
lighted
exam-
ples
In terms of the SIMANTICS platform:
• New release of VTT’s Apros ver. 6 (software for modelling and dynamic
simulation of processes and power plants) built on top of SIMANTICS 1.6.
• Integration between the Siemens Comos engineering system and the
Apros simulator via SIMANTICS.
• Integration of the Intergraph SmartPlant design system and the Apros
simulator.
• Integration between an automation CAD program and the Apros simulator.
• Integration of modelling and simulation capabilities utilising Modelica
simulation language, especially for mechanical engineering purposes, to
SIMANTICS.
• Design and implementation of a multibody system (MBS) modelling and
simulation environment demonstrator. The demonstrator utilises brings
together the 3D geometry modelling and visualisation capabilities of CAD
software and the OpenModelica language.
• Integration of a best in class life-cycle analysis (LCA) software, KCL-Eco,
to SIMANTICS. Due Thanks to SIMANTICS, additional capabilities are
enabled, e.g., connections to process simulators, interfaces to industrial
design systems and databases, possibility to use user-interfaces of other
engineering tools, and multi-user support.
In terms of virtual design environment:
• Integration of model checking to SIMANTICS. A more efficient and reliable
model checking enabled together with ability to verify the correctness of
larger systems than before.
• Implementation of the whole design chain of mobile machine cabins in
comprehensive virtual environment; regarding visual appearance, thermal
conditions, and sound or vibration experiences.
• Modelling, simulation and control of the entire exhaust tube of a combus-
tion-engine as a noise or vibration source, in all audible frequencies to
manage better the noise properties and user experience.
People
& net-
works
• The most significant industrial partners in the programme have been For-
tum, Wärtsilä, and Metso. In joint projects the sphere of partners consists
of dozens of firms.
• The open source SIMANTICS platform is today hosted by Simantics Divi-
sion of the THTH Association of Decentralized Information Management for
Industry with 25 industrial and academic members .
• Linkoping University and Modelica Association (Simantics, mechanical
engineering)
• Royal Institute of Technology (machine modelling)
• Luleå University of Technology (automation, Artemis activities, ProcessIT.
EU –strategy)
• Steering Group of the programme: Rauno Heinonen, Risto Kuivanen, Tuomo
Niskanen
• Core team of the programme: Olli Ventä (programme manager), Ismo Ves-
sonen, Riikka Virkkunen, Tommi Karhela, Timo Määttä, Teijo Salmi

eEngineering 2009—2012. Digitising the product process

Title eEngineering 2009—2012. Digitising the product process
Author(s) Kaisa Belloni and Olli Ventä (Eds.)
Abstract In addition to industrial production, the success of Finnish industry is
based strongly on the design and engineering of devices, working
machines, manufacturing plants, power plants, process machinery and
ships for global markets. At the same time, digitisation has become ever
more vital to the success of industrial production and engineering and
the volume, value and importance of the digital, virtual realm is increasing
dramatically compared to physical plants and machines.
At the beginning of the 2010s traditional heavy industry accounted
for 75% of the total value of Finnish exports, up notably from 57% in
2000. To reduce the design and production ramp-up times by half, VTT’s
eEngineering spearhead programme (2009-2012) developed technology
platforms for modelling and simulation, design knowledge management,
life-cycle management, and human-technology interaction. The high-
lights of the research carried out during the programme are presented
in this publication.
The most signiﬁcant achievement of the programme is Simantics,
an extensive operating system providing an open, high-level application
platform on which different computational tools can be easily integrated
to form a common environment for modelling and simulation. Programme
also enabled successful integration of user’s sound and noise experience
and thermal comfort modelling to the design of machine cabins in a vir-
tual design environment.
ISBN, ISSN ISBN 978-951-38-8125-2 (print)
ISBN 978-951-38-8126-9 (online)
ISSN-L 2242-1173
Date 2013
Language English
Pages 59 p.
Keywords System modelling, simulation platform, model integration, user experi-
ence design, machine cabin, auralization, virtual simulation environ-
ment
Publisher VTT Technical Research Centre of Finland
P.O. Box 1000
FI-02044 VTT, Finland
Tel. +358 20 722 111
Series title and number
VTT Research Highlights 8

eEngineering 2009—2012. Digitising the product process

More Related Content

What's hot (20)

Viewers also liked (13)

Similar to eEngineering 2009—2012. Digitising the product process (20)

More from VTT Technical Research Centre of Finland Ltd (20)

Recently uploaded (20)

eEngineering 2009—2012. Digitising the product process