Simulation Driven Product Development.

Simulation Driven Product Development
How it can be combined with Lean Philosophy to achieve increased

product development efficiency.
Master of Science Thesis in the Master Degree Program Supply Chain

Management
SARA JOHANSSON
DAVID SÄTTERMAN
Department of Technology Management and Economics

Division of Logistics and Transportation
CHALMERS UNIVERSITY OF TECHNOLOGY
Gothenburg, Sweden, 2012
Report No. E2011:021
How it can be combined with Lean Philosophy to achieve increased
product development efficiency.
Master of Science Thesis
SARA JOHANSSON
Master Degree Program Production Engineering
KTH ROYAL INSTITUTE OF TECHNOLOGY
DAVID SÄTTERMAN
Master Degree Program Supply Chain Management
Supervisors
Per Johansson
University supervisor
Department of Production Engineering
KTH ROYAL INSTITUTE OF TECHNOLOGY
Kent R. Johansson
Corporate supervisor
RTPM - Product Modeling Methods
SCANIA CV AB
© SARA JOHANSSON & DAVID SÄTTERMAN, 2012
Report No. E 2011:021

Department of Technology Management and Economics
SE – 412 96 Göteborg
Sweden
Telephone +46 (0)31-772 1000
Chalmers Edition
Reproservice, Chalmers
Gothenburg, Sweden, 2012

ABSTRACT
During the last decades the product lifecycle has been shortened and as a consequence the majority
of manufacturing companies have focused on decrease the time to market. Several companies within
the automotive industry have during the last years implemented Simulation Driven Product
Development as a method to reduce lead time, increase quality and reduce development cost for
their products.
This study is divided into three areas; (1) Investigate the applicability of Simulation Driven Product
Development in a Lean Product Development environment. (2) Examine the preconditions for
implementing Simulation Driven Product Development. (3) Study the possible effects in a product
development process of using simulations performed by design engineers.
By studying available literature in the area of Lean and Simulation Driven Product Development the
consistence of the two product development philosophies are rather high. Simulation Driven Product
Development tends to support the principles of Lean as well as reduce different kinds of waste that
normally is present at a R&D department.
In order to implement a Simulation Driven Product Development a number of preconditions were
identified; e.g. knowledge and experience within both physical and virtual testing, cross functional
synchronizations, trust for simulation results and incentives from management. These factors, among
a couple of others, are according to the theory and case studies preconditions for a successful
implementation where the primary case is based on Scania, but Saab Automobile AB, Jaguar Land
Rover and AB Volvo is also analyzed in the benchmarking study.
Simulations performed by design engineers will increase the potential of Simulation Driven Product
Development in several ways. A major finding from the study is that it enhances the understanding
for properties of the items among design engineers. The increased understanding also increase the
knowledge and decrease the learning cycle time for the design engineers. It also seems to be a
resource efficient way since waiting times are reduced. The need of support from simulation groups
might also reduce the communication barriers between different groups and increase the cross
functional work.
Keywords: Simulation driven product development, Simulation driven design, Lean Product
Development, Virtual testing, Automotive, Lead time, Scania, knowledge, Product Development
Process.
SAMMANFATTNING
Eftersom produktlivscykeln blir allt kortare har utvecklingstiden för produkter fått ett ökat fokus.
Inom fordonsindustrin har simuleringsdriven produktutveckling använts som en metod för att
reducera ledtid samtidigt som kvaliteten förstärks och den totala utvecklingskostnaden sänks.
Detta examensarbete är uppdelat i tre olika områden: Det första utreder huruvida simuleringsdriven
produktutveckling är förenligt med Lean produktutveckling. Det andra området fokuserar på
förutsättningar för att kunna implementera en simuleringsdriven utvecklingsprocess. Det tredje och
sista området fokuserar på konstruktörssimuleringars inverkan på produktutvecklingsprocessens
effektivitet.
Förenligheten mellan Simuleringsdriven produktutveckling och Lean produktutveckling är hög, enligt
befintlig litteratur. Simuleringsdriven produktutveckling stödjer flertalet av principerna till Lean
samtidigt som det bidrar till att reducera slöseri på forskning och utvecklingsavdelningen.
De två andra forskningsfrågorna angrips med hjälp av litteratur såväl som fallstudier. Den
huvudsakliga fallstudien genomförs på Scania, men även Saab Automobile AB, Jaguar Land Rover och
AB Volvo studeras i syfte att öka examensarbetets generaliserbarhet. Studierna identifierar ett antal
viktiga faktorer för att kunna implementera en simuleringsdriven produktutveckling: incitament för
en förändring från ledningen, kunskap inom både fysisk och virtuell provning, tvärfunktionell
synkronisering, en tydlig och framtung produktutvecklingsprocess och tillit till simuleringar.
Konstruktörssimuleringar kan på flera sätt höja potentialen hos simuleringsdriven produktutveckling.
Det ger konstruktören en ökad förståelse för sin komponents fysiska egenskaper. Tät feedback från
simuleringarna kortar inlärningstiden hos konstruktörerna och resulterar även i mer genomarbetade
produkter. För att lyckas med dessa simuleringar krävs ett nära samarbete mellan konstruktion och
beräkning vilket bidrar till att öka det tvärfunktionella arbetet och minskar
kommunikationsbarriärerna mellan konstruktions- och beräkningsgrupper.
.
PREFACE
This Master of Science thesis was conducted during the spring in 2012 on behalf of Scania CV AB. The
study was a collaboration between the master programs Production Engineering at KTH Royal
Institute of Technology and Supply Chain Management at Chalmers University of Technology.
We would like to thank our supervisor at Scania CV AB, Kent R Johansson for giving us the
opportunity to conduct this study and for his support and continuous feedback on our work. His
belief in this subject and our study has helped us getting valuable input and strengthen the
importance of the research. We would also like to thank all the members of the steering group, Emil
Axelson, Magnus Bergman, Hans Holmlöv, Patrik Karlsson and Johan Tingström, since they have has
been a great resource for us to gain a broader understanding and a way to discuss our theories and
results.
Further on, we would like to thank our supervisor from KTH, Per Johansson, who has helped us to
assure the academic content and the scientific contributions. Thanks are also directed to Ola
Hultkrantz, examiner at Chalmers, and Per Johansson, at KTH, for their approval to a collaboration
between the two universities and that we have had the possibility to conduct the thesis together. It
has been very valuable for us to address this study with two different academic backgrounds.
We would also give thanks to Tomas Sjödin, former employee at Saab Automobile AB, Jesper Blixt at
AB Volvo, Lars Tengelin and Neville Johnson at Dassault Systèmes for their contribution to a broader
perspective of this subject outside of Scania.
Finally, thanks to all the employees at Scania CV AB that have participated in the interviews and have
taken their time to give an honest imagination of the current situation and contributed with
suggestions for improvement. It is because of the open minded employees at Scania we fell proud to
present this thesis.
This study has given us great insight about simulations within product development, but also about
Scania and their organization and it will have a great value for both of us in the future.
Södertälje, May 2012
Sara Johansson David Sätterman

CONTENT
1 Introduction__________________________________________________________________ 1
1.1 Background ______________________________________________________________ 1
1.2 Purpose _________________________________________________________________ 2
1.3 Problem Analysis __________________________________________________________ 2
1.4 Delimitations _____________________________________________________________ 3
1.5 Outline of the Thesis _______________________________________________________ 3
2 Methodology _________________________________________________________________ 5
2.1 Method Model of the Study _________________________________________________ 5
2.1.1 Method to find output to RQ1 ___________________________________________ 6
2.1.2 Method to find output to RQ2 and RQ3 ____________________________________ 6
2.2 Data Collection ___________________________________________________________ 6
2.2.1 Document Study ______________________________________________________ 6
2.2.2 Observations _________________________________________________________ 7
2.2.3 Interviews ___________________________________________________________ 7
2.3 Reliability and Validity ______________________________________________________ 8
3 Theoretical Framework _________________________________________________________ 9
3.1 Product Development in General _____________________________________________ 9
3.1.1 Traditional Product Development Processes and Methodologies ________________ 9
3.1.2 Different Roles in the Product Development Process. ________________________ 10
3.2 Lean Product Development _________________________________________________ 11
3.2.1 Knowledge-Value_____________________________________________________ 15
3.2.2 Definitions of Waste __________________________________________________ 18
3.2.3 Set-Based Concurrent Engineering _______________________________________ 22
3.2.4 Cost Awareness ______________________________________________________ 24
3.2.5 Cadence and Synchronization ___________________________________________ 25
3.3 Simulation Driven Product Development ______________________________________ 26
3.3.1 Simulation Methods __________________________________________________ 27
3.3.2 Economic Aspects of an implementation __________________________________ 34
3.3.3 General Preconditions for an implementation ______________________________ 35
3.3.4 Plausible Difficulties in an implementation_________________________________ 37
3.4 Scania Product Development _______________________________________________ 37
3.4.1 The Product Development Process of Scania _______________________________ 38
3.4.2 Description of R&D Factory _____________________________________________ 40
3.4.3 GEO _______________________________________________________________ 44
4 Results From Case Studies ______________________________________________________ 47
4.1 Scania__________________________________________________________________ 47
4.1.1 Interaction between Simulation and Design Groups _________________________ 47
4.1.2 Interaction between Simulation and Physical Testing Groups __________________ 50
4.1.3 The Role of the Design Engineer _________________________________________ 51
4.1.4 Simulation and Analysis executed by the Design Engineer _____________________ 55
4.1.5 Synchronization and Geometry Assurance at Scania R&D _____________________ 56
4.2 Benchmarking of Simulation Driven Product Development ________________________ 56
4.2.1 Cost Rationalization by using Simulations at Saab Automobile AB _______________ 57
4.2.2 Towards Zero Prototyping at Jaguar Land Rover ____________________________ 60
4.2.3 Simulation Development Loops at AB Volvo ________________________________ 61
5 Analysis and Discussion ________________________________________________________ 65
5.1 Feasibility of combining Lean and Simulation Driven Product Development ___________ 65
5.1.1 The impact on the Lean Product Development Principles. _____________________ 65
5.1.2 The impact on Waste in the Product Development Process____________________ 66
5.1.3 Flow and Cadence ____________________________________________________ 68
5.1.4 Set-Based Concurrent Engineering _______________________________________ 69
5.1.5 The impact on Knowledge Creation ______________________________________ 69
5.2 Preconditions for Simulation Driven Product Development at Scania ________________ 70
5.2.1 Knowledge and Experience _____________________________________________ 70
5.2.2 Concurrent Engineering________________________________________________ 71
5.2.3 Simulations in the Product Development Process ___________________________ 72
5.2.4 Incentives for an Implementation ________________________________________ 73
5.2.5 Trust to Simulation Results _____________________________________________ 74
6 Conclusions _________________________________________________________________ 77
6.1 Result of the Research Questions ____________________________________________ 77
6.2 Recommendations to Scania ________________________________________________ 80
7 Further Studies ______________________________________________________________ 81
References ______________________________________________________________________ 83
1 INTRODUCTION
This chapter provides a background to the subject of this master thesis project. It also conducts the problem
analysis and delimitations as well as an outline of the thesis.
1.1 Background
The product lifecycle has during the last decades been shortened, therefore time to market has been
more crucial for all parties related to the product development process (Johannesson et al., 2005).
Due to the increased focus on a shortened development phase of products, several companies have
put effort to create efficient product development processes. Lead time is not the only important
parameter; quality and cost are examples of other parameters that can be influenced by the product
development process.
From a quality perspective the focus is not only to perform the tasks in an accurate way, it is also
important to focus on the right tasks (Bergman & Klefsjö, 2010). In today’s organizations it is
common that they try to build in quality into the system. The Lean enterprise, originated from
Toyota, uses the Japanese word Jidoka to describe how quality is achieved in production (Liker,
2004). Jidoka contains several different activities; e.g. automatic stops, error proofing, and in-station
quality control. When Swedish enterprises talk about Lean, the Jidoka uses to be replaced by the
expression “right from me”, which also relates to build in quality into the system.
Another cornerstone of the Lean philosophy is the waste reduction; to achieve a Lean production
several different types of waste needs to be minimized. The waste reduction in combination with the
built in quality and continuous improvements is an important part of generating a high performing
production site. A lot of attention is put into the production process and create efficient flow in
production. However, the Lean enterprise has potential in several different areas outside production,
e.g. product development and services (Liker & Morgan, 2006). Waste is not only present among
blue-collar workers, it is likely that the waste awareness is at least equally important in the area of
white-collar workers as well.
By studying the learning process of an engineer, trial-and-error cycles are important for the learning
outcome (Jacobs & Herbig, 1998). An engineer in the field of mechanical engineering for instance
needs to test a solution in order to get a feeling for how to handle strength behavior of different
geometries. The learning-by-doing process gives the engineer the opportunity to enhance its skills
over time and is a part of creating a learning organization. It is very seldom that the first solution to a
problem or a task is the right one. Almost all solution is a result of an iterative process with a mixture
of synthesis and analysis (Johannesson et al., 2005). Nowadays, Computer Aided Design, CAD, in
combination with simulation tools can be used to create virtual models of the item. By using virtual
model instead of physical prototypes the lead time has a potential of being shortened significantly
(Johannesson et al., 2005).
1
CHAPTER 1 - INTRODUCTION
The shortened lifecycle in combination with increased cost awareness and high demand of quality
are driving forces to develop an efficient product development process. Both the philosophy of the
Lean enterprise and simulation tools has potential to achieve increased efficiency in the product
development processes. However, by combining them the product development process can
probably be even more efficient.
Scania CV AB is one of the world leading manufacturers of heavy vehicles and marine and industry
engines. The usage of its modular product system allows it to offer a big variation of products and a
high degree of customization while keeping down the cost to its customers (Scania CV AB, 2011). To
be as efficient as possible, Scania continually needs to improve their methods and processes to keep
being profitable and competitive.
One of the most important demands for Scania’s products is dynamic strength durability to achieve
as high quality as possible. It is therefore a big focus on testing with physical prototypes in rigs and
on test tracks in their product development process. To use physical prototypes is both expensive
and time demanding. One way to become more effective is to reduce the numbers of physical test in
the development projects. At the same time Scania aiming to achieve superior quality on their
products and reduced lead time for the development projects.
1.2 Purpose
The purpose of this study is to investigate the current product development process at Scania and
find benefits and drawbacks in the existing structure of the process regarding simulations and lean
product development. The study also investigates how Lean principles and simulation driven product
development can be integrated and what the limitations can be with these methods.
1.3 Problem Analysis

One of the biggest wastes for the R&D department of Scania is likely the rework made due to bad
and uncertain designs of the components. The root cause can be the lack of knowledge when the
design engineer needs to take decision about concepts in an early phase. This uncertainty can often
be avoided by using an iterative simulation driven product development process.
Based on the background, problem analysis and purpose, three research questions are formulated in
order to be answered in the result chapter of the thesis.
RQ1: How can iterative simulation driven product development be integrated with Lean product
development concepts?
RQ2: What are the preconditions for a simulation driven product development?
RQ3: In what way can simulations performed by the design engineers improve simulation driven
product development?
2
1.4 Delimitations
The study focuses on the mechanical engineering track of the product development process at Scania
CV’s R&D department in Södertälje, Sweden. The insights and potentials in the embedded system
and software engineering processes will therefore be delimited. The reasoning of choosing the focus
on the mechanical track is related to the university study background of the authors.
The study will primarily have the design engineers’ point of view and investigate the need for
simulation tools in the product development process. How engineers in the technical simulations
process and the testing processes can use iterative simulation will not be investigated since this
processes can be looked upon as supporting processes to the design process. Therefore will this
study mainly involve interviews with design engineers, though both simulation engineers and
engineers within physical testing will be involved.
The simulations that will be investigated in this study are mainly calculation focused simulations,
especially Finite Element Analysis. Simulations within digital manufacturing and other areas outside
of the Research and Development organization at Scania are excluded.
The study will investigate preconditions for implementing simulation driven product development. A
Yes-or-No question regarding if simulation driven product development should be implemented or
not is not the focus of the study since it already seems obvious to recommend it. Though, advantages
and disadvantages regarding simulation driven product development will be analyzed and discussed.
1.5 Outline of the Thesis

This section comprises an overview of the structure and information of content for each chapter
headline in the thesis
1. INTRODUCTION
This chapter gives an academic and corporate background to the subject of the study as well as
clarifies the research questions who are supposed to be answered by the study. It is also aiming to
emphasize why the topic of the study is of interest.
2. METHODOLOGY
This chapter provides the different methodologies used to achieve the output of the three research
questions. It also comprises a reliability and validity discussion of the study.
3. THEORETICAL FRAMEWORK
This chapter provide a deeper understanding for the topic of the thesis. It gives an overview of Lean
Product Development and Simulation Driven Product Development as well as introducing Scania
specific theory in order to give an enhanced understanding for the specific circumstances for the
study.
5. RESULTS FROM CASESTUDIES

This chapter is divided into two different subchapters. The first subchapter provides an case study
used to analyze the current situation of Scania regarding the processes and usage of simulations. The
second subchapter provides case studies from three other companies within the automotive
industry, which is aiming to increase the generalizability of the study.
3
5. ANALYSIS AND DISCUSSION

This chapter concludes the analysis and striving to come up with output to the research questions.
Findings from the theoretical framework and the results from case studies are used as basis for the
analysis. The first section of the analysis has a general approach and investigates primarily RQ1 while
the second part is more business specific and focus on RQ2 and RQ3.
6. CONCLUSIONS
This chapter comprises the result of the study. Each one of the three research questions are
answered as well as it provides some specific recommendations for Scania of how to proceed with a
simulation driven product development approach.
7. FURTHER STUDIES
This final chapter aims to provide recommendations of areas feasible for further studies where the
thesis lack in depth or areas which has not been included in the study.
4
2 METHODOLOGY
This chapter is going to highlight the different research methods that will be used during the project and connect
each method to academic theories.
2.1 Method Model of the Study

This master thesis project is performed at the R&D department of Scania CV AB and emphasize the
possibilities of implementing a simulation driven product development process. In addition to Scania,
other companies implementation of simulation tools are being studied both through primary data,
i.e. data collected in the study through interviews and written correspondence, and secondary data,
i.e. data and information collected by someone else accessible through articles and conference
material (Björklund & Paulsson, 2007). Due to the company focus, the research design of the study is
a case study. Davidsson and Patel (2003) emphasize that the case design is useful when studying a
process and possibilities of change, which is analogous to the objective of this study. By using a case
study approach it is possible to use several different methods to collect data such as interviews,
document and observations (Yin, 2008). Wallen (1996) emphasizes that a case study gives the
possibility to gain knowledge and great depth understanding for the focal company.
A case study has limitations and drawbacks as well, Wallen (1996) highlights that it usually is very
time consuming and it might be difficult to finish the study. Another disadvantage is that if the study
is too closely connected to a certain case it is a risk to lose its general applicability. To handle the
time issue, the delimitations are a key by delimiting the scope of the outcome within appropriate
time. The theory framework as well as benchmarking cases from other companies give the scope of
this thesis a broader perspective than it would had if it only would have been a case study of Scania.
The method model, see Figure 1, describes the procedure of how data and knowledge is gained in
order to answer the three research questions; RQ1, RQ2 and RQ3.
Figure 1 Method model for the Study
5
CHAPTER 2 - METODOLOGY
The procedure to generate the output of the research questions differ some depending on the
questions, therefore the method is presented individually in following sections.
2.1.1 Method to find output to RQ1

RQ1 is not company specific and has more of an academic approach than the other research
questions. Therefore, the answer to the questions is found by a literature review, which is carried out
to develop the theoretical framework. A literature review is a cheap and rather quick method to
gather information for a study (Björklund & Paulsson, 2007). To get an enhanced understanding for
Lean Product Development, several different kinds of sources is used describing the philosophy in
different ways. It is easy to find sources about this subject so the challenging part is to filter what
sources that actually contributes to the framework. Since the research within simulation driven
product development is limited, it is more difficult to find accurate literature about the area.
Therefore the subchapter about this field in the literature review is more related to best practice
examples than academic reports. A simulation driven product development’s impact on the values of
Lean product development are then analyzed in order to find out whether the Lean and Simulation
Driven Product Development is consistent or not.
2.1.2 Method to find output to RQ2 and RQ3

RQ2 and RQ3 share the same methodology to receive the output to the questions. The study has its
origin in the theory chapter. Based on the theory, conclusions from the specific areas presented in
the results from case studies chapter are drawn. This kind of research there theory is used as a
starting point for principles followed by conclusions for a specific case is known as a deductive
research approach (Davidsson & Patel, 2003). The results are based on the case study concluding
qualitative interviews with in total 63 participants. A qualitative study has the advantage of providing
as depth of understanding for the specific situation (Holme & Krohn, 1997). Another advantage is the
flexibility of the method; it is easier to adapt a qualitative study during a project than to adapt a
quantitative. However, a qualitative research might be influenced by the interpretations when
transferring information between the interviewee and the researcher (Bryman & Bell, 2011). When
using qualitative research instead of quantitative it is also an increased risk of decreasing the study’s
generalizability (Björklund & Paulsson, 2007).
2.2 Data Collection

According to Andersen (1994), there are three ways to collect data: (1) Document studies, (2)
Observations and (3) Interviews. Which method that should be used should depend on the purpose
of the investigation, the case background, the length of the study and available resources.
2.2.1 Document Study

The purpose of the document study is to use data that is already spoken, written or published
(Andersen, 1994), e.g. literature, annual reports and articles. The documents used in this study are
mainly books and articles from journals, but master thesis reports, webpage information, conference
documentation and internal documentation at Scania is also used when there is not any book or
journal article in the specific subject. According to Andersen (1994) more general documents should
be used in the beginning of the study to than narrow the subject and read about more specific areas.
Therefore are books about lean in general first read to than later narrow the subject for the
literature, i.e. journal articles about simulations within product development and lean product
development.
6
CHAPTER 2 - METHODOLOGY
2.2.2 Observations
Through observations, either participating observation or non-participating observation, the social
behaviors can be seen directly (Andersen, 1994). The researcher is always in direct contact to the
objects been observed and observations are preferably used in the beginning of the study to create a
more clear view of the problem (Andersen, 1994).
This study includes different types of smaller observations, all non-participating, to get a better
overview and understanding for different areas regarding simulation and development projects. The
observations sometimes lead to some supplementary questions in order to deepen and secure data
from the observed situation. Examples of observations are pilot projects regarding simulation
performed by design engineers, interference analysis secession and project planning meetings.
2.2.3 Interviews
An interview can be performed either verbally or written through a survey (Andersen, 1994). All
interviews for this study are verbally while some follow-up questions are answered through written
correspondence.
To gather information about the product development process at Scania, interviews are made in
three rounds; Pre-study, first round and second round. Interviews can be divided into structured,
semi-structured and unstructured (Björklund & Paulsson, 2007). The pre-study interviews are made
with a selected few people to get an overall picture of the process and most of this interviews are
unstructured since the authors has not knowledge enough at this time to define areas and relevant
questions. The first round of interviews is then with many more people in different areas at R&D to
get a both broader and deeper perspective. Here are the interviews semi-structured since a semi-
structured interview has the discussion area predefined and the interviewer has prepared some open
questions (Björklund & Paulsson, 2007). All the semi-structured interviews in this study are recorded
and one person will lead the interview and ask questions while the other interviewer will take notes
and make sure that no questions are forgotten.
The last round of interviews are more of a discussion based type where different conclusion made
from the first round are discussed and analyzed together with mostly the same persons from the pre-
study interview round. These interviews are sometimes unstructured and sometimes semi-structured
depending on the subject.
There are also interviews and reference visits hold at other companies than Scania. This interviews
and visits are made after the first round of interview at Scania so a comparison can be made more
easily and what is relevant to study at the other companies is clearer. The reference visits are made
within the automobile industry since it is an industry that has applied simulation driven product
development and Lean product development in a larger scale and have used it for a longer period of
time. For Saab Automobile AB and AB Volvo, the interviews are made directly with the employees
while for Jaguar Land Rover, they are held through Dassault Systèmes.
All interview rounds are qualitative and not quantitative since knowledge and understanding about
the product development process are more important than a large number of data. Although some
quantitative questions might be used during the first round if there is a desire to compare different
departments internally or with external companies regarding working hours or economical aspects of
product development.
7
CHAPTER 2 - METODOLOGY
2.3 Reliability and Validity

Regarding validity, it is divided into internal and external validity (Bryman & Bell, 2011). The external
validity is how much the study can be generalized (Bryman & Bell, 2011) and the internal validity is
about the methods used to really measure what was intended to measure (Gripenberg, 2006) and
this can be achieved by triangulation (Bryman & Bell, 2011).
Since this study does not only contain the case study at Scania, but also a literature review and
benchmarking at other companies the external validity is fairly high. The internal validity is ensured
by continuously having a dialog with both the supervisor and the steering group members about
methods used for the study and by letting them read both the theoretical framework and results
from case studies during the study to cover up gaps or correct misunderstandings.
According to Bryman and Bell (2011) the reliability is whether a study is repeatable or not. The
reliability to this study does depend on the way correspondents to the interviews are chosen and
how the questions and answers can be interpreted. When correspondents are chosen, a conscious
the selection should be made regarding what would give the best result (Creswell, 2003). In this
study, all the correspondents chose to participate or not and when choosing persons to the
interviews, it will be employees with a good analytic skills and an aptitude to express opinions be
preferred.
The language used at the interviews at Scania does include internal terms. When using these terms in
the thesis, first translated to a more general language and than from Swedish to English it might be
more difficult to interpret. Therefore the reliability of this thesis is also dependent on how well the
internal terms are described.
8
3 THEORETICAL FRAMEWORK
This chapter is aiming to give a deeper understanding for the subject and current research. The chapter starts by
describing traditional product development and Lean product development. Next section describes simulation
driven product development and its precondition. It also comprises a briefly introduction to different simulation
methods. Finally the chapter introduces the specific circumstances for product development at Scania CV AB.
3.1 Product Development in General

The product development process at different companies has developed during the last century. In
the beginning of the 20th century it used to being more of single craft man role to develop a product
from scratch until it reach fulfill the needs of the customer. Nowadays product development to
comprise large complex organizations there several individuals are experts in a more specific field.
3.1.1 Traditional Product Development Processes and Methodologies

According to Ward et al. (1995), the traditional U.S. product development process is influenced by
Shingley’s book Mechanical Engineering Design. Shigley (1963) suggest an approach based of
stepwise iterations. The engineer should start by understanding a problem and when the problem is
understood a solution is generated. Based on next step the solutions is adapted to fulfill the targets it
did not reach in the first run, see Figure 2. Even though the upcoming phases are not involved in the
previous steps, Shigley puts effort to the importance of making decision based on knowledge.
Shigley’s approach use to be named Point-Based Concurrent Engineering (Ward et al., 1995).
Figure 2 Example of Point-based engineering approach
Johannesson et al. (2004) introduce some different methods for development processes common in
Swedish industries. One example of a process is the axiomatic product development process. It
contains four different sub-processes; (1) A problem definition process in which the customers’ needs
are defined and transformed to functional requirements and constraints. (2) A creative synthesize
process where design parameters are generated in order to fulfill the functional requirements from
previous phase. (3) An analysis process there the different alternatives are evaluated by simulations.
(4) A validating process there the final solution is verified to fulfill all requirements in the
specification. Johannesson et al describe a stepwise approach where different phases are quite
separated from each other and there is a lack of cross-functional work, except for phase shifts when
the knowledge is transferred between departments.
9
CHAPTER 3 – THEORETICAL FRAMEWORK
Another process described by Johannesson et al is the systematic product development process. The
first step in this approach is the product specification phase in which the specification and needs for
the product are created. Next phase is the concept generating phase. In this step, several possible
solutions are combined and compared in order to find the best one. A concept is often chosen from
matrices where the designer compares benefits and drawbacks of each one to finally choose one
concept. Since the cost for changes is low in the start of a project it is important to evaluate the
concepts early (Johannesson et al., 2004). The concepts are evaluated with the specification in mind
and have to fulfill the requirements to not be eliminated. The design engineers tend to do the
concept generation and elimination with only limited support of cross functional competence. Firstly
after the concept generation phase has come to its end the cross functional functions are involved in
order to improve and validate the chosen concept, see Figure 3.
Figure 3 Illustration of traditional Swedish approach of product development processes
Johannesson et al. emphasizes the importance of frontload decisions in the processes due to
increasing cost along the timeline. Both processes described above do suggest an approach where
the concept is chosen primarily on the basis of the design engineers knowledge and skills.
3.1.2 Different Roles in the Product Development Process.

The engineer role was founded during the industrialization, in the 19th century (Johannesson et al.,
2004). Before the industrialization took place it was all about handcraft. One single person used to be
responsible for the styling, design and manufacturing for a product. During the industrialization, the
styling and design became separated from the manufacturing and the engineer as a profession took
place. However, even though the engineer has become more specialized than what craft work used
to be, it had a wide spectra of tasks and responsibilities. On other areas, the same individual was in
charge to secure the supply of raw material and choosing manufacturing techniques. The origin in
craft also made the design of the products quite artistic; one example is the old steam locomotives
from the 19th century that is both functional and very esthetic (Johannesson et al., 2004).
10
During the 20th century the role of the engineer has changed (Johannesson et al., 2004). Since the
development project has grown in range it is a need for more specialized competence, see Figure 4.
Examples of new roles connected to engineering were for instance; design engineer, simulation
engineer, testing engineer, styling engineer, material expert, production engineers, purchaser and
project manager.
Figure 4 The new roles of engineering
The design engineer has focus on the design and it needs to interact with all other functions. The
simulation engineer has together with the testing engineers often a shared responsibility to validate
the durability of the design. The styling engineer focus on the esthetic part of the design, they often
also have the responsibility for the ergonomics. Material engineers have core competence in material
properties, it is often a research function that supports the design engineers with feasible
alternatives when it comes to the choice of raw material. The production engineers are responsible
for the manufacturability of the design and how it will be manufactured. The purchaser source raw
material as well as have the contact with sub-contractors when outsourcing is used. Since the
projects have become more complex and involve lots of different functions, a coordinator is needed;
this is the role of the project manager.
The roles in the product development process has changed during the last centuries, from a
individual job with a wide spectra of different kinds of responsibilities, to complex team activities
where each individual’s core competence is needed in order to success with a development project.
3.2 Lean Product Development

The Lean enterprise originates from Toyota Motor Company, but the term Lean was first introduced
in 1991 in the book “The machine that changed the world” (Morgan & Liker, 2006). In this book, Jim
Womack and Daniel Jones (1991) used the term Lean manufacturing as doing more of everything
with less of everything. The Lean manufacturing system was described as a better, faster and
cheaper production system with less inventory and labor hours required (Womack & Jones, 1991).
Womack and Jones are though fast to point out that Lean manufacturing is just one part in the Lean
enterprise, which also includes marketing, distribution, accounting and product development.
Although, most companies has just focused on the manufacturing part of Lean (Morgan & Liker,
11
2006). It is a good first step, but according to Morgan and Liker (2006), you need to take the second
step upstream to product development in order to become a Lean enterprise. However, in the book
about Lean product development at Harley-Davidson, “The Lean Machine”, Womack points out that
you should not bring Lean manufacturing upstream to product development, (Oosterwal, 2010).
Womack emphasize that the Lean aspects within manufacturing and product development are very
different and even though they might seem similar at first sight he recommends that an expert in
Lean manufacturing should not give advice within Lean product development (Oosterwal, 2010).
According to Allen Ward, Lean product development is about knowledge-based product
development which has its origin from Orville and Wilbur Wright (Oosterwal, 2010). The dream
about flying was an obsession among humans during the 1900’s and Ward therefore saw it as
astonishing when the first ones creating a flying machine operated by humans did not come from a
university, government or big company, but from two uneducated brothers. Ward emphasize that
the Wright brothers did not just invented the first airplane, but they also invented a new way of
developing products. They created a wind tunnel to test different air foils in order to create trade-off
curves and gain knowledge about aerodynamics (Oosterwal, 2010). Instead of first designing an idea
and then test it, the Wright brothers first gained knowledge about wing profiles, power application
and dynamics of the airplane which they used to design a possible solution. This approach resulted in
a process that took them about $1000, five years and a powered flight at the third attempt
(Oosterwal, 2010). According to Ward the Wright brothers’ method was then used during World War
II to create airplanes with a very short lead time and those engineers later ended up at Toyota Motor
Company after the war ended (Oosterwal, 2010)
Liker and Morgan (2006) wanted to find the answer to what the underlying principles of product
development were that had made Toyota so successful. They identified 13 principles built on three
subsystems: (1) Process, (2) Skilled People and (3) Tools and Technology, see Figure 5.
Figure 5 Coherent systems approach to product development (Morgan & Liker, 2006)
These three subsystems are further defined with 13 principles, see the box below.
12
THE 13 PRINCIPLES OF LEAN PRODUCT DEVELOPMEMT SYSTEM

(MORGAN & LIKER, 2006)
The Process Subsystem: Includes all the activities and the order of the activities required
to bring a product from idea to start of production. Starts with the raw material,
information of customer needs, competitive product data and engineering principles,
which is being transformed to the engineering of a product built by manufacturing.
• Principle 1: Establish Customer-Defined Value to Separate Value-Added Activity from

Waste. Define your customer is always the starting point in a Lean system. Than
define what your customer wants. Everything that is not of value for your customer is
waste. In product development, waste is divided in to two broad categories: Waste
created by poor engineering and Waste in the product development process itself.
• Principle 2: Front-Load the Product Development Process While There Is Maximum

Design Space to Explore Alternative Solutions Thoroughly. The earlier several
alternatives are explored, the more opportunities are still possible. Therefore, Toyota
puts most of the work and effort as early as possible, also known as front-load, in its
projects. The problem solving is made with several possible designs and by using this
set-based (parallel solutions) approach, it increases its chance of finding the optimal
solution.
• Principle 3: Create a Leveled Product Development Flow. Once value is defined, waste
can be reduced from the process in order to achieve a waste-free Lean Product
Development System. A Lean Product Development System is a knowledge work job
shop which can be continuously improved by eliminating waste to level workload,
synchronize cross-functional processes and to reduce rework to a minimum.
• Principle 4: Utilize Rigorous Standardization to Reduce Variation, Create Flexibility and

Predictable Outcomes. There are three types of standardizations used by Toyota: (1)
Design standardization; achieved through common architecture, modularity and
shared components. (2) Process standardization; made by designing components
based on a standard manufacturing process. (3) Engineering skill set standardization;
used to have a flexibility in their staff and project planning.
The People Subsystem: This subsystem covers recruiting, training engineers, leadership,
structure of the organization and the culture including language, symbols, beliefs and
values.
• Principle 5: Develop a Chief Engineer System to Integrate Development from Start to

Finish. The Chief engineers role is to be the project manager, the leader and technical
system integrator. Since the chief engineer has the full responsibility of the project,
this person knows exactly the status of the project as it also knows all the technical
details.
• Principle 6: Organize to Balance Functional Expertise and Cross-Functional

Integration. This principle is how to find the balance between functional expertise
combined with a seamless integration between the different disciplines. Toyota is
functional organized and is integrated through the chief engineers, module
development teams and an obeya (big room) system, where cross-functional
meetings are held.
13
• Principle 7: Develop Towering Technical Competence in All Engineers. Since the

automobile is such a complex system it is of great interest to have a lot of technical
excellence in many different areas, like aero and fluid dynamics, mechanics,
electronics and computer technology. Although many automotive manufactures put
more effort into broaden their engineers rather than encourage them to deepen in
their subject. Internal courses and training can sometimes be so general that it value
can be questioned. At Toyota the focus lies in learning more core engineering in their
own area, rather than getting a MBA.
• Principle 8: Fully Integrate Suppliers into the Product Development System. Since
suppliers usually provide more than 50 % of the vehicle, they need to be a part of the
Lean Product Development System. The same way a company manage and nurture
their own manufacturing and resources should the suppliers being managed and
nurtured. Suppliers are involved form the earliest stages in the development process
and provide Toyota with guest-engineers to work full-time at Toyota in order to
integrate them more with the company.
• Principle 9: Build in Learning and Continuous Improvement. The most sustainable

advantage a company can have is the ability throughout the organization to learn and
improve.
• Principle 10: Build a Culture to Support Excellence and Relentless Improvement. Since
the core values are seen in the Toyota work, the culture supports the principles and
makes it a living part of how Toyota gets things done.
The Tools and Technology Subsystem: This subsystem consists of the CAD systems, digital
manufacturing and machine technology as well as the “soft” tools used for problem
solving, learning and best practice.
• Principle 11: Adapt Technology to fit your People and Processes. Adding new
technology tools must be made so it becomes integrated with the already existing
process or it will probably retard the work rather than help. Toyota works hard to
adapt its design software and digital simulation to the Toyota way.
• Principle 12: Align your Organization through Simple, Visual Communication. One
main tool used is hoshin kanri, known as breaking down big corporate goals into
meaningful objectives at working level. It is also used to break down vehicle
objectives down into system objectives for performance, weight, cost and safety.
Toyota also use simple, visual methods for communications like A3-sheet reports with
only the most important, knowledge-based information on them.
• Principle 13: Use Powerful Tools for Standardization and Organizational Learning. In
order to work with kaizen, continuous improvement, standardization is required. This
occurs on all levels from the overall design process down to individual engineering
check-lists.
The principles should be seen as points to discuss rather than rules and they would not explain how
Lean product development works in reality. To learn the principles and the Lean product
development process should be done step by step, like peeling of one layer at the time from an
onion (Morgan & Liker, 2006). The more layers you had peeled away, the sooner you will discover
the basis of both Lean manufacturing and Lean product development: the importance of integrating
people, processes, tool and technology to add value to the customer and society (Morgan & Liker,
2006).
14
3.2.1 Knowledge-Value
Compared to Lean production, Lean product development is focused more on knowledge flow than
the flow of physical objects. While manufacturing creates the actual products, development creates
the operations how to manufacture the products and usable knowledge. Therefore the biggest value
adding activity in development is to create new knowledge(Ward, 2007). According to Ward(2002)
missing or unused knowledge is the reason to project failure and knowledge is created by three types
of learning:
• Integration learning is about your customers, supplier, manufacturing, the product usage and
environment, etc. From here the knowledge can be found of how to develop regarding to
other criteria and the surrounding environment.
• Innovation learning is the knowledge of how to create new concepts and possible solutions
to problems.
• Feasibility learning is about taking the right decisions and chose the best concept or solution
regarding to cost, quality and time.
Ward mentions that useful knowledge from both the integration learning and feasibility learning is a
critical part of the development process since it is here you can gain time and effort by avoiding
mistakes (Ward, 2007). By turning data into readable and understandable information the results
from analysis and test can be more widely spread and understood.
One way to transforming data to be more easily understood is to use trade-off curves for different
product preferences (Ward, 2007). They are used to see how close to the limit you are by letting a
parameter differ (Holmdahl, 2010) and these curves should be the center of the development
activities (Ward, 2002). One example could be a diagram, depending on a screw dimension, with
curves showing how it will affect bearing stress, sheer stress and tensile stress, see Figure 6. By
displaying it as curves instead of tables makes it easier to reuse the knowledge gained from the test
by others not familiar with the original test. The strength with the curves is how easy it is to see if
you are beneath a safety zone or not, by drawing red lines, limit curves, for the mandatory criteria
(Holmdahl, 2010), see Figure 6.
Figure 6 Example of trade-off curves
15
Since the methods today are more specialized, engineers are either a design engineer working with
Computer-Aided Design, CAD, or a calculation engineer working with Computer-Aided Engineering,
CAE, i.e. Finite Element Method, FEM and Computer Fluid Dynamics, CFD (Holmdahl, 2010). It is
more important to be able to understand the different needs. However, the role of the design
engineer might change over time since different CAE tools are becoming more integrated with the
CAD-programs. Many software providers believe that the next generation of design engineer does,
except for specific product areas, need more insight in CAE tools (Honeywill, 2008).
Ward has two learning cycles which can be used to create knowledge. One is the value-adding cycle,
see Figure 7, and the other on is the LAMDA™-cycle, see Figure 8. The cycles are similar in many
areas and both of them handle how to handle problems and gain knowledge from them.
teach observe
Go see
connect
to invent
theory
(In) form Ask why
abstract test
Figure 7 The value-adding cycle (Ward, 2007)
Go see can also be referred to as Genchi Genbutsu, go and see or go to the gemba. Literally Genchi
Genbutsu means to focus on the actual part, the actual place (Morgan & Liker, 2006). It means that
you should go to the place (or person), the gemba, where the problem or work actually happens to
understand with own eyes the current reality (Liker, 2004). This activity is called Look in the
LAMDA™ cycle (Ward, 2002), see Figure 8. For a design engineer the go and see activity can be to
walk through the assembly line, talk with a customer or see an experiment (Ward, 2007). Many
engineers today dedicate their working hours to either meetings in conference rooms or working at
their desk, very few get to see the actual products, especially while it becomes more common with
outsourcing manufacturing and virtual engineering without any physical prototypes (Morgan & Liker,
2006). One important Genchi Genbutsu-activity, according to Liker and Morgan, practiced by Toyota
is the prototype building phase (2006). Early in the project the engineers themselves get to mount on
their prototypes to the vehicle in order to get knowledge about the surroundings and get a feeling of
their own design.
16
Ask why, also referred to as The Five Why:s or Ask in the LAMDA™ cycle, is meant to find the root
cause for different problems (Liker, 2004). According to Liker, Toyota asks Why five times to go down
all the way from there the problem started in order to get rid of the real problem and does not just
do a quick-fix solution. Ones you reached the root cause, it is time to find a solution for it. This can be
done by the next step, (In)Form or Model-Discuss-Act, see Figure 7 and Figure 8. The “Five why”
process has unfortunately become a “Five Who” process of blame at many companies, including
Harley-Davidson and it is important that the learning process stays around the “why” and not the
“who” in order to be effective (Oosterwal, 2010).
The (In)Form part in the cycle is there we form the information gained from the other steps (Go see
and Ask why) into something usable like a drawing, prototype, report or equation. These are also
used to inform others (Ward, 2007). According to Ward, many organizations fail with the value-
adding cycle for two reasons. The first reason is that not all steps in the cycle are performed, or
sometimes not even aware of, all steps of the outer cycle, see Figure 7 . The second reason is that
you are taught in school that you learn by repeating the teacher and very few lessons is about
inventing something new yourself. We learn that scientific and mathematical principles are already
known and there is nothing more to discover, but since all the competitors also know Newton’s laws,
engineers have to learn how to discover and learn new knowledge in order to gain competitive
advantages (Ward, 2007). Therefore Lean companies know that they have to teach the engineers
how to learn besides the traditional educational system and that is the primary source of Lean
competitive advantage (Ward, 2007).
Look
Act Ask
Discuss Model
Figure 8 LAMDA™ cycle (Ward, 2002)
Model in the LAMDA™ cycle can be the same as the form in the value-adding cycle. The important
thing is to create easily understandable models that are clear and visual (Ward, 2002). Discuss can be
replaced with the Inform activity which has its purpose to spread the knowledge and get opinions
from many different areas. When all the information has been found, analyzed and discussed, you
can finally Act and implement a solution for the problem (Ward, 2002).
17
3.2.2 Definitions of Waste

In order to work with continuous improvement you need to define value and waste, so waste can be
eliminated (Morgan & Liker, 2006). Ward (2007) categorizes the activities in product development
into three different kinds; (1) value adding, (2) non-value adding but necessary and (3) waste. If
activities should be classified as value adding or not depends on what the process is supposed to
deliver. It might be a drawing, a solution or knowledge to the operation. What waste is, depends on
the expected outcome, the company and the department. For product development, waste has been
explained by both Ward (2007) and Morgan and Liker (2006) in different ways.
3.2.2.1 Waste according to Morgan and Liker
Morgan and Liker (2006) define waste in product development as any activity in a process that needs
resources without adding value for the customer. They have divided waste activities into seven
categories, see Table 1. These categories are originally mentioned in the Toyota Way (Liker, 2004) for
Lean production, but with some changes it can be used for product development as well (Morgan &
Liker, 2006).
Table 1 Seven categories of waste (Morgan & Liker, 2006)
Example in product
Waste category What does it mean? Example in manufacturing
development
Unsynchronized concurrent Producing ahead what is

Producing more or earlier
Overproducing tasks, too much information in actually needed by the next
than the next process needs
the report step or the customer
Operators waiting for
Waiting for material, Waiting for simulation results,
Waiting material, waiting while an
information or decisions prototypes or decisions
automatic machine is running
Moving material or Unnecessary hand-offs,
Moving parts or products
Conveyance information from place to information changing hands by
unnecessary long distances
place word, pictures or data
Engineering errors,
Doing unnecessary
reinvention, simulation on the Unnecessary or incorrect
Processing processing on a task or
wrong model – lack of processing
doing an pointless task
standardization
Direct result of
A build-up of material, Information waiting in queues
overproduction. Having more
Inventory models or information that to be processed, hides
in stock than the required
is not being used problems in the information
minimum for a pull system
Long travel distances, Unnecessary movement for
Excess motion or activity
Motion unnecessary meetings and the operators. Reaching or
during task execution
project reviews looking for tools and parts
Inspection to catch quality
Rework, late changes, excess
Correction problems or fixing an error Inspection, rework and scrap
tool tryout
already made
18
The goal for Lean is not only to eliminate waste, muda, but to eliminate the three dimensions of
waste, muda, mura and muri (Morgan & Liker, 2006).
THE THREE M:S (MORGAN & LIKER, 2006)

Muda (non-value added): Includes all the seven wastes mentioned in Table 1. Any activity
that add cost or lengthen the lead time to a product without support from the customers
need is should be considered as muda.
Muri (overburden): When a machine, process or person is pushed to work beyond it

natural limits. Can cause quality problems and safety risks which will lead to rework.
Mura (unevenness): Sometimes it is too much work for the engineers right before
deadlines, which usually follows by a calm period with little to do. An uneven workflow
will create muda and muri.
3.2.2.2 Ward’s Definition of Product Development Waste

Ward (2007) has a different definition of waste in product development compared to other studied
authors. The focus is on waste of knowledge, where Ward uses three different categories to describe
it, see Figure 9.
Figure 9 Knowledge waste in product development processes adapted from Ward (2007)
Waste caused by Scatter

Scatter is common in the processes in many different ways. It can be from a time perspective;
employees’ time is scattered among different tasks and a set up time is required each time an
employee is interrupted. According to Ward (2007), it is a common contra productive behavior when
a project is delayed. Imagine a design engineer waiting for a prototype, when it is worried about the
delayed delivery it calls the purchasing department several times a week. Due to that, the purchasers
are interrupted in their work by answering calls from worrying engineers and they cannot focus on
their main task. Ward highlights two main root cause categories for scatter; Communication barriers
& Poor tools.
19
Communication barriers
Lack of communication is a common source of scatter waste. It might be physical barriers that make
it impossible to share knowledge, e.g. incompatible computer software or departments located far
away from each other. Another issue in communication is the social barriers, these might be caused
by lack of vertical and/or horizontal integration within the organization. However, too many
meetings will keep the engineers busy and slow down the process (Ward, 2007). Knowledge barriers
drives communication issues, if the receiver does not have sufficient knowledge to understand the
information it receives, the message is lost and the knowledge will be wasted. A fourth category of
communication barriers is the information channels, it is related to how information is distributed
and stored in the organization.
Poor tools
If the process is too detailed, the engineers are forced to use tools that are not sufficient for their
certain situation (Ward, 2007). However, according to Ward standardization should be used to create
a robust process. The same task should be performed in the same way every time. But the focus
should be on how to transfer data to knowledge, not a detailed step-by-step instruction.
Waste caused by Hand-Off
This category is about separating knowledge, responsibility, feedback and action from each other
(Ward, 2007). The effect of Hand-off is that decision making is made by people that either have a lack
of knowledge in the subject or not having the authority to carry out the decisions (Ward, 2007). An
example of hand-off is when companies divide the responsibility for the engineering design among
CAD-operators and analysts. Ward introduces two different root cause categories for Hand-off;
useless information and Waiting.
Useless information
A lot of time in a developer’s working day is used to pursue useful information among useless
information, a result of scatter. Sometimes useless information is related to overproduction. Some
engineers tend to get stuck in endless optimization for a project or designing new products when the
old ones were good enough to keep.
Waiting
In production, the sequence thinking is obvious an item that flows through the processes. However,
Ward (2007) emphasizes that the sequence thinking is not suitable for a product development
environment. In product development the next process has to influence the previous one to create a
successful solution. Ward suggest cross functional teams from the earliest phase in the projects to
create an enhanced understanding and broader perspective of possible solutions. Involving an
analysts too late in the project might cause a lot of rework.
20
Waste caused by Wishful thinking

Many developers come up with a specification in an early phase of a development project. However,
due to development lead times it is uncertainties in the customer demand and often the customer
does not really know what it wants at this stage (Ward, 2007). Actually, the design decisions are then
based on speculation and guesses instead of data and facts.
Some design engineers tend also to stick with a concept or solution from a very early stage. If this
happens it is impossible to ensure that the solution actually is the best possible solution to fulfill
what the customers really wants (Ward, 2007). Wishful thinking is derived to two root causes;
Testing to Specification and Discarded Knowledge
Testing to Specification.
Ward (2007) doubts the behavior of testing to specification. A test cannot verify that a product is
ready for the market, it is impossible to do enough tests to check e.g. fatigue issues. When U.S. car
manufacturers were benchmarked to Toyota, Ward found out that Toyota uses 70-90 percent less
prototypes. However, Toyota still outperforms the U.S. manufacturing regarding quality.
Testing is necessary during the product development process, but should be used in the right way.
According to Ward it should be used to gain the knowledge in order to avoid traps. Testing engineers
and simulation engineers should come up with the answer and knowledge of why a concept breaks
down and suggest improvements to the design engineer even if it actually fulfills the specification.
Ward suggest two different kinds of testing. The first category is tear down testing where the test
runs until the item breaks in order to gain knowledge, notice that the test should not be stopped
when specification is fulfilled. The second one is validating testing in the final state of the
development process in order to validate the concept as well as the manufacturing process.
Discarded Knowledge
According to Ward (2007), it is crucial for the knowledge development how knowledge is handled
and created during a project. Focus is often on the lead time and getting the product to market, not
how to sustain the knowledge. If more attention would be put on sustain gained knowledge, the lead
time could be shortened next time a similar project is in progress. Tests, performed with focus on
specifications, make it difficult to reuse the gained knowledge in another project. Data and
information that comes up during the project have to be transferred into knowledge, the data in
itself does not develop excellent engineers (Ward, 2007).
3.2.2.3 Scania R&D Factory’s Definition of Waste
In the compendium R&D factory Scania introduces guidelines for how to make decisions towards a
robust product development process (Scania CV AB, 2010). Different kinds of waste in the Product
Development process are defined, these are presented in the box below. Scania CV AB (2010)
identifies seven different possible types of waste at the R&D department:
21
DIFFERENT CATEGORIES OF WASTE IN PRODUCT DEVELOPMENT

ADAPTED FROM SCANIA CV AB (2010)
1. Doing more than necessary: It can for instance be the creation of documentation with
more content and information than necessary, or order more prototypes than actually
needed for a test. This kind of waste can also be present in meetings if more people than
actually need to participate at a meeting are invited or the more PowerPoint slides than
necessary are prepared for a presentation.
2. Not helping others If you have finished your part of a project you should help others
instead of try to optimize a solution more than good enough. To not support others is a
waste of your own resource and knowledge.
3. Not finding necessary information easily: When it is a lack of accessibility to information

a waste occur. It can for instance be a homepage that is difficult to navigate or not divide
responsibility clear enough. It can also be related to excess scattering of information;
instead of only send an email to the individual that need the information the information
is sent to everyone
4. Missing or unclear decisions: To not respect taken decisions about deliveries cause
waste. It can also be a lack of decision in the cadence, the manager might be unavailable
when a top down decision is needed, which cause uncertainty.
5. Excessive lead times in the planning stage: Excessive lead times tend to hide issues and
cause a false sense of security. It also causes waiting time in the process due to that an
item, which drawings are ready for simulation or testing too early, has to wait for the next
step.
6. Having excessive wait: This kind of waste is related to both waiting for physical
prototypes such as waiting for late people attending a meeting. Excessive waiting cause
idling and unused resources. It is often related to disturbances in one way or another.
7. Not using everyone’s competence: Every individual’s competence should be used to gain
the performance.
These seven categories are introduces in the folder as examples of waste, however, it is a possibility
to have more categories of waste in the process.
3.2.3 Set-Based Concurrent Engineering

Set-Based Concurrent Engineering, SBCE, is an important part of Lean product development and is
described as a cornerstone and success factor in the philosophy (Ward, 2007; Kennedy,
2003).According to Morgan and Liker (2006) SBCE is a part of the second Lean product development
principle; Front-Load the Product Development Process While There Is Maximum Design Space to
Explore Alternative Solutions Thoroughly. The target is to avoid early development decisions that
cause late engineering quick fixes to cope with the demanded market launch. These quick fixes
cannot be viewed as continuous improvement, actually, they should be treated as waste (Morgan &
Liker, 2006). The basis from SBCE is that all cross functional stakeholders of the new product starts
generating alternative solutions by their own. These alternatives are combined into possible sets of
solutions (Ward, 2007). The solutions are eliminated one by one based on facts from, for instance,
simulation and analysis related to shortcomings in the solutions. However, parallel solutions are
carried along almost the entire product development process. When the narrowing process is
22
finished the overall best solution is left and the decision is knowledge based see Figure 10 (Sobek et
al., 1999).
Figure 10 Concept narrowing in Set-Based Concurrent Engineering
3.2.3.1 Framework for Set-Based Concurrent Engineering

Three of the four fathers of the terminology Set-Based Concurrent Engineering, Sobek, Allen Ward
and Liker launch in the article Toyota’s Principles of Set-Based Concurrent Engineering a framework
for SBCE. Since they are the owner of the expression SBCE (Morgan & Liker, 2006), the framework
can be viewed as it is representative for SBCE in general. Sobek et al. (1999) identifies three
principles as a basis for the SBCE; (1) Map the design space, (2) Integrate by intersection and (3)
Establish feasibility before commitment. These principles together with some implementation
principles for each one of them creates a framework for working with several concepts in parallel and
knit them together, the framework is described in detail in Appendix B.
Sobek et al. (1999) emphasize that it is impossible to do a step-by-step guide how to success with
SBCE. As much of the Lean Enterprise, it is all about culture. The tools might be helpful, but to
success with Lean a pure cultural change has to occur (Liker, 2004). Therefore, the success factor of a
SBCE implementation is more about the implementation of the ideas than the principles themselves
(Sobek et al., 1999).
3.2.3.2 The Economics of Set-Based Concurrent Engineering
The main idea with SBCE is based on the fact that the cost for a change increases for each step closer
to market launch a project takes (Johannesson et al., 2004). By maximize the learning in the early
phase and minimizing late changes, it is a potential of saving a lot of money (Ward, 2002). SBCE is
more or less a methodology aiming to decrease risk and increase quality. In almost all projects, it is a
risk of a concept failure where rework is necessary. By base the decision on facts and investigate
several different solutions, this risk is minimized and the probability of rework and failure decreases
exponentially when the number of possible solutions increases (Reinerstein, 2009).
23
However, it is a tradeoff between minimize the risk of rework and cost for investigating several
solution paths. Increased number of investigated subsystem increases costs for simulation, testing
and physical prototypes. Reinerstein (2009) introduced an adapted variant of the Wilson formula,
also known as Economic Order Quantity, EOQ, formula, aiming to give the “break even” for number
of solutions versus risk of failure. Reinterstein highlights the importance of cost awareness among
engineers using SBCE in order to optimize the profit.
3.2.4 Cost Awareness

Usually, continuous improvement and kaizen is focused on enhanced quality or reduced lead time,
but when Toyota faced a great loss caused by the oil crises and great recession in 2009 they started
to work with Kaizen from an economic perspective (Liker, 2011). In production, different costs were
tracked on billboards and all effort was put in how to reduce scrap and use junk to rebuild the plants.
Some plants and production lines had to be rebuilding in order to become more flexible and produce
smaller vehicles. Carts and racks that would normally have been purchased were built in-house by
scrap metal instead. Old overhead catwalks were made into carts. One production line budget on
originally $4 million was trimmed down to $700 000 (Liker, 2011).
A common issue in product development is the lack of cost awareness at the R&D department
(Reinerstein, 2009). The management has often clear targets regarding cost and several KPI:s are
related to cost, but the engineers, who actually are the ones that have the direct impact on the
expenses, does not bather of cost at all. Instead the design engineers tend to focus on proxy
variables, e.g. robustness, functionality and style. If the economic connection is absent, the design
engineer does not manage to understand its Cost of Delay, COD (Reinerstein, 2009). Due to the lack
of insight in COD, the prioritization between different projects is based on feeling rather than fact.
If the duration and COD is known for a project, Reinersten (2009) suggests that a weighted factor
should be used in order to evaluate the order of project execution. If the COD is divided by duration,
the project with the largest fraction should be given the highest prioritization. Table 2 gives an
example of a fraction calculation of three different projects.
Table 2 Example of prioritization by a weighted value
𝐂𝐎𝐃
Project number Duration Cost of delay Weight =
𝐃𝐮𝐫𝐚𝐭𝐢𝐨𝐧
1 1 3 3
2 5 9 1,8
3 10 27 2,7
To illustrate the impact project implementation order has on cost, the generated cost is visible in
Figure 11. In this example, it is assumed that all three projects will be delayed with at least its
duration time.
24
Cost Affect of Prioritization

600
Cost [Units]
400
200
0
1,2,3 2,3,1 3,2,1 3,1,2 1,3,2 2,1,3
Order of finnishing project
Project number 1 Project number 2 Project number 3
Figure 11 Example of the impact on cost due to implementation order
According to both the graph in Figure 11 and the fraction guide in Table 2, the most optimal project
implementation order is project 1-3-2. In this certain example the cost would have been reduced by
10 % if the design engineer uses the weighted fraction instead of following the project number, i.e.
order: 1-2-3.
To be able to do this kind of prioritization the knowledge of COD for each project is essential.
Therefore, it is of importance for the design engineers to get an enhanced understanding for COD of
different projects they are involved in.
3.2.5 Cadence and Synchronization

Cadence initiates the pulse to the process. When it is present in the process, activities occur at a
regular basis and it is possible to foresee and plan when certain activities are going to be executed
(Reinerstein, 2009). Synchronization is more about integration. In a synchronized process activities
are planned cross functionally and different departments gets a common milestone (Reinerstein,
2009). It is possible to have a cadence in the absence of synchronization, for instance if each
department has its own cadence and plan the projects independent of other departments. It is also
possible to have synchronization without cadence; product launches can for instance be executed
continuously, but without a predicted frequency. Since the time between two launch dates might
vary cadence is not present.
Cadence is one of four concepts that the management of Toyota product development rely on
(Ward, 2007). It is crucial to level the workload in the process as well as secure that internal
deliveries in a project not are delayed. Cadence also makes it possible to plan the project more
efficient since the supporting activities of a process are predictable. Milestones can be used to
sustain the cadence and the number of milestones is crucial for the efficiency of the process. Too few
gateways will cause waste in the process, due to large batches of information and waiting time
(Reinerstein, 2009). If large batches of information are delivered to the process all at once, the
feedback will be delayed. As a consequence, iteration loops and rework will be time consuming.
Therefore are several short meetings is a better approach than a few long meetings to create an
efficient process. Cadence can be initiated to the process in several ways with different targets, see
Table 3 (Reinerstein, 2009).
25
Table 3 Different kinds of cadence adapted from Reinerstein (2009).
Type of Cadence Description Result
Projects cascading into each other

New projects are launched at a
Product Introducing Cadence can be eliminated and create a
regular basis.
leveled work load.
Create leveled work load for the
The deadline for submitting items testing department as well as
Prototyping Cycles to a test is fixed and testing occur minimize the number of tests since
at a regular basis projects can be coordinated
regarding the test cycle.
When a the portfolio is updated
The product portfolio is updated
regularly it is possible to ensure
New Product Portfolio Screening regularly, for instance daily or
that all involved parties have the
weekly
same version of an item
Information delivered in small
Reviewing design at a regular
batches make it easier for more
Design reviews frequency, do not wait until the
rapid feedback and find problems
drawing is ready
proactively
Ward (2007) proposes a project oriented cadence. The demand time and product launches are the
crucial parts for finding the correct rhythm. The synchronization or milestones can be used in
combination with cadence to sustain small batches and minimize queues in the process.
3.3 Simulation Driven Product Development

Simulation Driven Product Development, SDPD, and Simulation Driven Design, SDD, has been
frequently discussed in the automotive industry for the last decade. For instance Volvo Cars
Corporation designed its S80 model by great support of simulations already 2001. This was the first
car where the strength the spot-welding was predicted by computerized calculation instead of
conventional testing (Westling, 2001).
The differences between SDPD and SDD are their implementation range. In SDD, simulation is the
basis when generating the design. The design engineers make, for instance, concept choices based
on simulation. In SDD, simulation still might be complementary to the physical testing which tends to
be the primary way of verifying a solution. In SDPD, simulations are the basis for the entire product
development process, not only decisions related to the design phase. It can for instance comprise
different kinds of virtual testing, virtual factories etc. Physical testing might still be necessary to, for
instance, verify legacy requirement and capture the reality in order to be able to generate accurate
simulation models. However, in a SPDP environment, physical testing is a support function to virtual
testing and not vice versa. By using simulation at an early phase of the product development process
companies has a potential of reducing time to market, increase quality and reduce costs (Adams,
2006). However, even though the methodology has been in use for more than ten years, it is difficult
to find academic publications of the topic. Most of accessible information is related to best practice
articles.
This section will give a theoretical basis for possible simulation methods to use as well as highlighting
some examples of documented preconditions for a successful implementation of SDPD. Finally, it
also comprises economically findings of simulation driven design implementations.
26
3.3.1 Simulation Methods

Different simulation methods are mainly used today by manufacturing companies to support the
product development process (Hirsch, 2007). Nowadays, the whole product development and
production cycle can be computerized, the three most important parts are: (1) Computer-aided
design, CAD; (2) Computer-aided engineering, CAE; and (3) Computer-aided Manufacturing, CAM
(Hirsch, 2007). These three areas create the different phases of a virtual prototyping environment,
see Figure 12.
Figure 12 The virtual prototyping environment adapted from Hirsch (2007)
In order to use any CAE tools, such as FEM or CFD, a CAD model needs to be created first. That model
will then be the input to the CAE simulations. The results from the CAE simulations will then be used
to redesign the previous CAD model. This is repeated in an iterative simulation driven design cycle
until the CAD model has the demanded functionality. When the design objectives are reached CAM
software is used to simulate the manufacturing processes, such as milling parameters and tool routes
(Hirsch, 2007).
Since the methods today are more specialized, engineers are either a design engineer working with
Computer-Aided Design, CAD, or a calculation engineer working with Computer-Aided Engineering,
CAE, i.e. Finite Element Method, FEM and Computer Fluid Dynamics, CFD (Holmdahl, 2010)..
However, the role of the design engineer might change over time since different CAE tools are
becoming more integrated with the CAD-programs. Many software providers believe that the next
generation of design engineer does, except for specific product areas, need more insight in CAE tools
(Honeywill, 2008).
27
CAE is mainly used in the automobile industry to decrease cost and time to market (Raphael & Smith,
2003) and some of the most common CAE tools and methods used within this industry are described
further below.
3.3.1.1 The Finite Element Method
The Finite Element Method, FEM, is a methodology developed in the 1950’s to analyze expected
durability of a physical geometry (Adams, 2006). When analyzing durability of different geometries,
partial differential equation is used to describe the geometry analytically (Hutton, 2005). However,
the partial differential equations tend to be complex and very difficult to solve analytically.
Therefore, numerical methods, such as FEM, are necessary; these methods generate a numerical
approximated solution to the equation. FEM nodes are combined into different kind of elements, see
Figure 13 (Hutton, 2005), which all are reasonable to describe analytically. By study each element,
the impact on the geometry as a unit can be approximated. The practical engineering application of
the mathematic methodology FEM is referred to as Finite Element Analysis, FEA.
Figure 13 Different types of elements used in the finite element method; U.L. Beam element defined of two
nodes, U.R. Triangular element defined of three nodes, L.L. Quadrant element defined of four nodes, L.R
Triangular element defined of six nodes.
The FEA is usually performed with separate software, like ABAQUS or ANSYS, but there are also
available FEM solvers on the market that are integrated with CAD software. The integrated
applications are usually considerably faster to use since many of the steps, like building up the mesh-
model, is made automatic. Though, the integrated application performs a simplified calculation and
cannot handle as complex systems as the separate programs. In comparison, both types of solvers
give a fairly similar result when it comes to where the critical areas are in the model, see Figure 14,
but the separate solvers are better to determine the absolute values of the tensions (Loman
Strinnholm, 2012).
28
Figure 14 Comparison between calculations models where the left model comes from ABAQUS and the right
from GAS in CATIA.
3.3.1.2 Crash Simulation

Computer simulation has had a large impact on crashworthiness of the design. It is possible to do
many more iterations with virtual crash methods than physical tests since a crash simulation goes a
lot faster and is cheaper than a physical crash test (Thomke & Fujimoto, 2000).
A crash simulation is based on a FEM simulation, but it still differs from static strength simulations
regarding the calculation methods. For static loads in FEM, implicit methods are used to invert the
matrix and find a reasonable solution. The crash simulations are instead calculated with explicit
methods since they need to take big deformations, plasticity and hardening factors of the materials
into consideration (Karlsson, 2006). The element grid has historically been coarser for crash
simulation models due to limits in computer capacity, but this is not always the case today because
of the evolution in computer performance over the last years.
The short period of time that is simulated affects the material properties and how fast the materials
stretch needs to be considered since the hardening varies with the stretching velocity. Therefore,
behaviors cannot as easily be intuitive predicted as static strength analysis or fatigue simulations. The
stiffness in the components are put against each other and fairly small changes in the strength in the
components can rise to completely new modes for indentations (Karlsson, 2006).
The crash simulation models also need to take the driver and passenger behaviors into consideration
which other simulations usually do not need. There are standardized models of drivers and
passengers today on the market and their behaviors are calculated either with FEA or rigid body
dynamics (Karlsson, 2006).
Almost all the simulation programs for FEM today have the possibility to do explicit calculation, but
some are more suitable and common for crash simulations like PAM-CRASH, ABAQUS and LS-DYNA.
The automotive industry is leading in this type of simulations and use crash simulations to reduce the
29
number of physical prototypes and crash tests in order to reduce both cost and time to market. One
development project at BMW contained 91 iterations on the design with help of crash simulations
and those iterations improved the crashworthiness by 30% for the side-impact, which would be very
difficult to accomplish with only physical crash tests (Thomke & Fujimoto, 2000). The physical tests
made in this project was also very time consuming and more expensive than the simulations, see
Table 4
Table 4 Approximate Lead time and Cost for a Development project at BMW AG, Germany (Thomke &
Fujimoto, 2000)
Problem-Solving Step Simulation (per iteration) Physical Prototype (per iteration)
Technical Meeting Planning and Piece part Design

Design
• <0.5 days • >2 weeks (involves many
meetings)
Data Preparation and Meshing
• Small change: <0.5 Design and Construction
days • Using existing model: 3 months
Build • Significant change: 1 ($150,000 per prototype)
week • New model: >6 months
• Entire automobile: 6 ($600,000 per prototype)
weeks
Crash Simulation
Crash Physical Prototype
• 1 day (varies with
Run • 1 week (includes preparation
computer hardware)
of test area)
$250 per day
Data Preparation and Analysis
Post-processing and Analysis • 1 day (crash sensor data only)
Analyze
• <0.5 days • 1-3 weeks (data, crash films
and analysis of physical parts)
Total time About 2.5 days to 6.3 weeks About 3.8 to 7 months
Total cost About < $5,000 About >$300,000
3.3.1.3 Computational Fluid Dynamics

Computational Fluid Dynamics, CFD, is a type of CAE simulation used to analyze fluid and thermal
transport. It can for example be used to analyze the flow through the different stages of fixed and
rotating blade rows in a turbine, compressor in an aircraft engine, the combustion process in diesel
engines and to predict the pressure drop in an exhaust system.
The governing equations that are used in CFD are based on three fundamental physical principles: (1)
conservation of mass, (2) conservation of momentum, and (3) conservation of energy (Anderson,
2009). This set of equations is known as the continuity equation, the Navier-Stokes equations, and
the energy equation.
30
THE CFD SIMULATION IN FIVE STEPS ADAPATED FROM HIRSCH (2007)

1. Select the mathematical model and define the level of approximation of the
reality that should be simulated.
2. Discretization of the space by define the grid generation and of the equations by
define the numerical scheme.
3. Analyze of the numerical scheme and establish its properties of stability and
accuracy.
4. Select an appropriate time integration method and a subsequent resolution

method in order to obtain the solution.
5. Post-process the numerical data graphically so the result can be understood and
interpreted easier, see Figure 15 .
Figure 15 A CFD analysis of a truck’s external flow.
CFD should be used in combination with pure theory and experiments about fluid dynamics since
they complement and support each other and CFD is therefore used for both basic mechanical
research and engineering design (Anderson, 2009). One example is the wind tunnel testing that has
been used to make the CFD methods and simulations even more accurate and it had decreased both
cost and time for especially developing new air craft designs (Anderson, 2009).
31
3.3.1.4 Interference Analysis

An interference analysis is performed inside a CAD program and is used to find interferences
between different geometries. The interference analysis can be divided between different
subsystems on the product to control where to look for interferences. It can also be regulating the
distance between two geometries that will count as interference, i.e. geometries which tangent with
each other, interferences for real, see Figure 16, or have a smaller gap between two surfaces than 2
mm. This type of simulation together with digital assembly differ from FEM and CFD, since they are
problems of fit and not function The answer to an interference analysis is either yes or no compared
to function problems, e.g. FEM, which have a more nuanced solution (Thomke & Fujimoto, 2000).
Figure 16 Interferences between two components from an interference analysis in CATIA
Boeing was one of the companies who started with interference analysis in the end of the 1990’s.
They build digital mock-ups of their 777 aircraft and used software to check for interferences during
a time period when the geometrical data was locked. The design engineers were informed by
interference for their components when the next design phase started which opened up for doing
design changes again (Thomke & Fujimoto, 2000). The interference checks found 251 interferences
that were prevented before they moved on the physical assembly tests. These interferences created
a stronger interaction between different design engineers since they needed to solve the
interference issues together (Thomke & Fujimoto, 2000).
32
3.3.1.5 Digital Assembly

Digital assembly is usually performed with applications integrated in CAD software, e.g. DELMIA in
CATIA. Digital assembly tools can identify risks in the manufacturing process in an earlier phase in the
development process, i.e. in the conceptual design phase (Diesel & Gas Turbine Publications , 2003).
Digital assembly is mainly used in the automotive industry where the time between the design phase
and start of production is critical. A digital assembly simulation can be used to decide the assembly
sequences for new products and discover interferences and deviations without physical prototypes
and assembly tools. If realistic manikins are used in the simulation, see Figure 17, ergonomics for the
operator and assembler can also be evaluated (Dassault Systèmes, 2012).
Figure 17 Digital assembly of a truck chassis with help of digital tools and a manikin.
Toyota Motor Company was one of the first users of digital assembly when it started with Visual and
Virtual Communication, V-comm, in the beginning of the 21st century (Normile, 2001). Digital
assembly was mainly used to be able to communicate between different factories overseas and to
get feedback on the design from floor workers early in the design phase without physical prototypes.
With help of the digital manufacturing program, which contained more than just digital assembly,
they cut their development time from 18 to 13 months (Normile, 2001). In the annual report from
Toyota in 2003 the mention digital engineering as an invaluable tool to reach a lead-time reduced by
more than half and they see the V-comm system as a way to enable more digital assembly (Toyota,
2003).
33
3.3.2 Economic Aspects of an implementation

In order to estimate the return of investment, ROI, for a Simulation Driven Product Development,
SDPD, project two parameters is required; (1) The cost for implementation of SDPD, both the
investments cost of hardware and software as well as education and training of employees. (2) A
monetary value of savings of implementing the project. One of the most obvious impacts of SDPD is
the decreased number physical prototypes (Ansys Inc., 2011). If the simulations can explain the real
phenomena all physical testing that is used today will not be needed in the future. However, physical
testing is still necessary for validation. Though, the number of iterative physical tests is going to be
reduced if a successful implementation of simulation driven design is done. According to Jackson
(2007) the best in class companies in the U.S. have reduced their prototypes from 4.6 generations of
prototypes to a mean of 3 generations. Ansys Inc. (2011) emphasizes that best in class companies
have reduced their need of physical testing from five-six trial-and-error experiments to two validating
test cycles. How large the actual monetary savings will be is related to both the type of business the
focal company is active in and how large it is (Jackson, 2007).
Aberdeen Group (2006) presents a study examining the experience of SDPD at 270 American
companies. Aberdeen compares how the best in class companies have implemented SDPD and what
kind of consequences it will have on the cost and time to market. The companies are divided into
different categories depending on product complexity, see Table 5.
Table 5 General Characteristics of Product Complexity. Source: (Aberdeen Group, 2006)
Product
Number of Parts Length of Development Cost to Build a Prototype
Complexity
Low Less than 50 Between a week and a year $7 600
Moderate Between 50 and 1000 Between a month and 5 years $58 000
High Between 50 and 10 000 Between 1 and 5 years $130 000
Very High Between 1 000 and 100 000 Between 1 and 20 years $1 200 000
Aberdeen Group highlights a difference of 1.6 prototype generations between best in class and
average companies which correspond to an average project saving of between $12 160 and
$1 900 000 for each project. The annual savings is then related to how many projects a company
completes during a year.
In order to estimate the ROI, Ansys Inc. (2011) has collected data from a couple of cases where
virtual prototyping and SDPD has been implemented. One example is the awarded Department of
Defense HPC Modernization Program – DEW in the U.S. Virtual prototyping has enabled initial savings
of $13.8 million (IDC, 2011). The expenses for the program are $812 million and the lower and upper
bound of the ROI is calculated to 678% respective 1292 % (Ansys Inc., 2011). This might be a best
practice example, but Ansys emphasizes that the ROI rate is quite high overall and a ROI of 300% is
reasonable for several businesses.
34
It is not only the costs of prototyping and physical testing that is affected by SDPD. Intangible
parameters such as increased product quality, goodwill and increased brand reputation is often also
affected (Ansys Inc., 2011). According to Jackson (2007) companies which have used simulation tools
at all stages of the product development process have experienced less quality complaints from
customers. The result at the field will not emerge momentary, but in time the warranty cost curve
will decrease significantly. The savings due to reduced warranty complains tend to largely exceed the
investment cost for the virtual prototyping projects.
Another parameter that is affected by the implementation of SDPD is the lead time. By reducing the
number of prototypes series and replace them with virtual prototyping, time to market will be
decreased. Depending of the product complexity the Aberdeen Group (2006) estimate the lead time
for a physical prototype testing cycle be from 13 to 99 days, see Table 6.
Table 6 Prototype Time per Product Complexity
Product complexity Number of parts Time to build prototypes
Low Less than 50 13 days
Moderate Between 50 and 1000 24 days
High Between 50 and 10 000 46 days
Very High Between 1 000 and 100 000 99 days
Since best in class companies have reduced the number of prototype series they have the possibility
of launching their low complexity projects 21 days before its competitors who has not implemented
SDPD to the same extent. For a very high complexity product the corresponding value is 158 days
(Aberdeen Group, 2006). An earlier market release will probably increase the market shares of a
mature market or the profit margins if the market is immature when the conventional release were
supposed to take place. The benefits of decreased lead time and shorten time to market are difficult
to put into monetary value. However, it will be an advantage and should increase the ROI rate.
3.3.3 General Preconditions for an implementation

One issue with Simulations Driven Product Development, SDPD, is that the competitors have access
to the same software vendors and therefore, the simulation applications themselves do not give any
competitive advantage (Adams, 2006). The main difference among a best in class company and a
laggard is instead related to how the tools and methodologies are implemented into the product
development process of a focal company (Jackson, 2007).
One significant difference between the very successful companies and the others is in which stage of
the development process simulation tools are used (Jackson, 2007). The product development
process can be divided into three different phases; Design phase/concept generating, Test
Phase/Validation Phase and Post design release phase. Best in class companies have front loaded
their simulations more than the others, see Figure 18.
35
Simulation Utilization
100
80
60
%
40
20
0
Design phase Test phase Post design release
Best in class Average Laggard
Figure 18 Comparison of simulation utilization adapted from Aberdeen Group (2006). Different companies
were asked when they use simulations in the product development process. E.g. All Best in class companies
uses simulations in the design phase, but less than 80 % of the laggard companies do.
All best performing companies among the 270 companies involved in the study, that Figure 18 is
based on, uses simulation in the design phase (Aberdeen Group, 2006). Overall, best in class have a
more frequent use of simulation in all phases. How simulation is used in the product development
process will influence the performance of the company. Aberdeen Group (2006) emphasizes that
CAD embedded simulation capabilities should be used in the design phase. Their main argument is
that the design engineer does not need to spend time learning new software from scratch that only
will be use sporadically. Best in class companies are 63% more likely than others to use the
embedded simulation systems in the design phase (Aberdeen Group, 2006). The familiar layout and
accessibility in embedded simulation capabilities lower the entry barrier of using simulations in the
design phase.
According to Jackson (2007), the structure of the training programs for employees has an impact on
the result of the implementation. Aberdeen Group (2006) emphasizes that the best methodology for
education seems to be a combination of specific examples and individual performed training
exercises. This is analogous to Adams (2006) who add a suggestion that the training should be
mandatory in order to keep a high knowledge level among the employees. Adams also highlights the
importance of choosing the right pilot projects to reach success. If the initial projects are a success it
will be spread within the company. However, if it becomes a failure, a widespread program will never
gain commitment. Therefore, projects that are reasonable or almost guaranteed to success with
should be used as examples in an early phase (Adams, 2006). It is tempting to use the software until
the limit of its full potential, but if these pilot projects fail in an early phase it is a risk that the
methods not only lose their credibility of advance projects, neither will they success to gain
acceptance for the more simple tasks they actually have documented power of perform efficiently.
The management has a crucial role during the implementation of a program towards SDPD.
Management’s commitment to a change program is fundamental to a success. However, the
visionaries at a company have to be continuously questioned in order to develop robust and bullet
proof analysis and this is also a task for the managers (Adams, 2006).
36
3.3.4 Plausible Difficulties in an implementation

There are some common challenges to overcome in order to be successful with an implementation
program towards SDPD. Aberdeen Group (2006) emphasizes lack of available time among design
engineers as the most common doubt in organizations. Design engineers are already choked as it is
today and therefore, they do not have any extra time for doing simulations. Jackson (2007) assign
this time limitation issue as a paradox; one of the main advantages of implementing SDPD is the
shortened time to market. By using SDPD each project should consume less time and therefore, the
time issue should not be a problem. According to Jackson, it is more a question of how to use the
resources in the most optimum way. Aberdeen Group (2006) present analogous data to Jackson’s
statement; none of the most common actions taken among companies that have implemented SDPD
are related to lack of time among design engineers. Therefore, it is likely that the anxiety related to
work burden is groundless or at least the problem will be so easy solved that no specific actions are
required.
Another issue is the available knowledge at the focal company. Independent if the company uses
CAD embedded simulation systems or separate simulation software, experts in each field are
required for success (Jackson, 2007). Several companies have a lack of this expertise when starting its
implementation. Aberdeen Group (2006) emphasizes that this problem is one of the trickiest to
overcome; four out of five of the most common actions to overcome the initial threshold are related
to this issue.
A third difficulty to handle is the physical test correlation. When striving to go all the way with an
implementing program towards SDPD it is not enough to achieve directional answers from the
simulations, the output from simulation has to be accurate (Jackson, 2007). To reach accuracy in
simulations, the model and method have to be frequently updated and improved continuously until
they actually describe the reality (Jackson, 2007). However, measurements and data have to be
gathered from physical testing. Therefore, high accuracy in physical testing is a precondition when
developing simulation models for successful virtual testing.
Adams (2006) highlights the quality assurance as another obstacle. When the product development
process relies on physical testing, simulations is often used as an additional assignment. Therefore,
the quality of the analysis and simulation phase is not crucial; the physical testing still has an accurate
quality assurance process. When eliminating assignments of the physical testing in favor of virtual
testing the quality assurance has to be built into the new virtual testing process. Otherwise Adams
(2006) emphasizes a risk of building in quality issues to the new process instead of reaching increased
quality, which often is a target when implementing a SDPD program.
3.4 Scania Product Development

Since the primary case of the thesis study is Scania, a basic theoretical understanding for the
company is required to understand the circumstances for the result of case studies, chapter 4. In this
section the product development process at Scania is explained followed by the department
guidelines presented in R&D factory. Finally a briefly introduction of some company specific methods
such as GEO is presented.
37
3.4.1 The Product Development Process of Scania

The product development, PD, process comprise three different phases or sub-processes; yellow
arrow, green arrow, and red arrow, see Figure 19 (Scania CV AB, 2011). Frontload
Figure 19 Illustration of the Product Development Process of Scania (Scania CV AB, 2011)
The PD process rests on some fundamental principles or statements connected to how Scania
develop products:
• Product ownership – The line is the owner of the product. The projects can never own a
product.
• Cross-functional and parallel – Product development is cross functional and it should
influence the process from the initiation of a new project to product follow up downstream
in the product flow.
• Uncertainties – When developing new products, uncertainties are necessary and have to be
handled, iterations will be something natural.
• Configuration – The project plan should be performed by a cross functional team and has
milestone as a basis.
3.4.1.1 Yellow Arrow – Pre Development

The pre-development sub-process comprises three different areas of responsibility; Research,
Technology development and Concept development.
Scania uses Research to gain knowledge of the technique of tomorrow and find areas in which core
competence should be developed to get a leading market position for the future (Scania CV AB,
2011). The knowledge gained by research is going to be applied in technology development.
In Technology development, new techniques are applied to improve the properties of the Scania’s
future products (Scania CV AB, 2011). The new technique is evaluated and knowledge of which
technique is suitable to fulfill the customers’ needs is delivered to the next yellow arrow activity
concept development.
38
In Concept development, the concept which will be the basis for the new components is generated.
All concept will be related to a demand, e.g. from a customer, legal requirement or a cost
rationalization.
A concept is finalized and ready for the next step when following paragraphs are fulfilled (Scania CV
AB, 2011):
1. Performance/Property objectives are described
2. Profitability analysis has been conducted
3. The concept is planable
4. The concept has been modularized
5. The concept has been cross functionally accepted
3.4.1.2 Green Arrow – Product Development
A project in this phase is characterized by an ability to be planned in detail. All deliveries have a high
delivery precision and a project at this stage will most likely in the nearest future be available for the
consumer (Scania CV AB, 2011). The main activities and resources within the green arrow are
allocated to product development projects. A product development project has a predicted market
launch date as well as due date for start of production. The line, which includes all delivery functions
of Scania, is responsible for delivering in time. The Project office supports the project with methods
and tools for project management, set milestones etc, but the line is responsible for the delivery
(Scania CV AB, 2011).
At an early phase of the Green-arrow process a project configuration is performed. The project will
be divided into single activities that are necessary to fulfill to reach expected benefits, application
targets and demand for the project. The planning is performed by a standardized method named
Scania Project Planning, SPP.
The development of the component is continuously in progress through the entire project. The target
with the green arrow phase is to adapt the concept chosen in the yellow arrow to its environment.
The project manager keeps the control of the project through a couple of milestones. When projects
reach Start of Production, SOP, a process verification is done and some adjustments might be
necessary before the project is ready for Start of Customer Order Production, SOPCOP.
Except product development projects this phase also comprises three other type of projects; Design
On Line, DOL, Special Order, S-Order, and Fit For Use, FFU.
3.4.1.3 Red Arrow – Product Follow Up
In the red arrow phase, projects already launched in SOPCOP are treated. The projects are
maintained and updated with the purpose of improve quality or decreasing costs (Scania CV AB,
2011). A component deviation that reaches the consumer will cause negative goodwill for the brand.
Lack of quality at the field often causes large costs and therefore has a direct impact on the
profitability of Scania.
39
3.4.2 Description of R&D Factory

R&D Factory is a set of guidelines used to support the decision making and create a stable and
reliable product development process in order to keep the market position of the company and
satisfy the customers’ needs (Scania CV AB, 2010). Analogous to the guidelines for production, the
concept is often described with the metaphor of a house with a ground, pillars and roof, see Figure
20. Even though it looks quite similar to Scania Production System, SPS, the R&D Factory is adapted
to suit the product development instead of the manufacturing environment.
Figure 20 R&D Factory described with the house metaphor (Scania CV AB, 2010)
The house includes three different categories of boxes; the core value, the principles and the
priorities of the product development department at Scania. It is surrounded by three important
success factors for the performance of the department.
3.4.2.1 The Core Values of Scania R&D
The core values customer first, respect for the individual and elimination of waste reflect the culture
of Scania and are always present together, see Figure 21 (Scania CV AB, 2010). Since all three core
values are equally important for the company, it is no prioritization among them.
Figure 21 The Core Values of Scania R&D
40
Customer first refers to the importance of always having the customers in focus. The product
development process has to be driven by the demand (Scania CV AB, 2010). At the R&D department
it is two types of customer in focus; it can be either the end costumer or an internal customer, e.g.
production. One key to create a consciousness of the customer’s need is to have a focus on that the
employees are doing the right activities, i.e. activities that create value for its customer.
Respect for the individual is aiming to use the knowledge and experience that is available among all
employees. It also refer to the possibility for individual development (Scania CV AB, 2010). Another
example of respect for the individual is to fulfill promises and deliver in time.
Elimination of waste is related to eliminating disturbances in the process. Scania defines waste as
anything that does not provide value or benefits for the end customer (Scania CV AB, 2010).
Examples of waste at Scania R&D department are described in section 3.2.2.3. Then waste is
eliminated the competitiveness of Scania can be increased by increased efficiency and reduced lead
times.
3.4.2.2 The Principles of Scania R&D
The R&D factory house contains four different principles; Normal situation, Right from me, Demand
driven and Continuous improvement, see Figure 22. These principles are aiming to sustain a common
way of thinking in every situation (Scania CV AB, 2010).
Figure 22 The principles of Scania R&D
Normal situation – flow orientation describes how activities should be performed. The flow
orientation secure that the deliveries continuously flow through the value chain and not are stuck in
queues (Scania CV AB, 2010). The R&D department has two different types of flow present; (1) The
information flow which delivers information to a specific project, for instance a stress target that is
fulfilled for a certain component and (2) the knowledge flow which contribute to a knowledge bank,
see Figure 23
41
Figure 23 Illustration of the two types of flows (Scania CV AB, 2010)

If a component is tested until the acceptance criterion is met, a delivery takes place to the project.
However, the test can be continued until the component breaks in order to learn how it breaks. This
gains knowledge and is a delivery to the knowledge bank. The Normal Situation is supported by six
sub processes; Standardized methods, Modularization, CEPPSS, Cross-functional and parallel,
Visualization and Balancing.
Demand driven output is the second principle of Scania R&D Factory. The demand will be used to
control the flow. To have a continuous feedback between different parts of the flow is crucial for this
principle. The delivery shall not take place before it is needed. The core of the method is to not
process more information than is needed downstream the chain (Scania CV AB, 2010). The deliveries
are going to be in small batches in order to sustain the internal customer’s current need. The small
batches have a potential of decreasing the throughput time due to reduced waiting time and more
rapid feedback. It is easier to detect a mistake earlier if only a small batch is delivered every time.
Right from me refer to that deliveries to internal customers are performed right from the beginning.
There are methods, tools and instructions which facilitate this principle. If the standardized method
is followed and the delivery still not is satisfying, it is a shared responsibility among all parties to
improve the method (Scania CV AB, 2010). The internal customer evaluates the quality of a delivery
and is responsible to give feedback to the previous step if the delivery does not fulfill the needs. A
rapid feedback is important in order to minimize the effect of a mistake. The one whose delivery
does not fulfill the requirements is responsible for doing the rework. If it is any uncertainties of who
is responsible for the lack of quality, the parties should cooperate in order to come up with a solution
(Scania CV AB, 2010).
Continuous improvement is described as the engine of R&D factory. The starting point for
improvements is the normal situation. Disturbances and deviation from the normal state is assigned
as waste and is a great opportunity for improvement. Within the continuous improvement principle
the normal situation is maintained, challenged and improved. The continuous improvement is a daily
work and based on that the cooperation with improvement works (Scania CV AB, 2010). It is also
something every employee will participate in since everyone’s competence and knowledge are going
to be utilized. The normal situation and standardized method is not only the starting point for
continuous improvements, it is also a requirement; if there are no common method among the
workers it is no “best known procedure” to improve. The improvement work has to be based on
facts, it is not sufficient to trust intuition for this kind of work.
42
3.4.2.3 The Priorities of Scania R&D Factory

Scania has four priorities included in the R&D factory; (1) Health and environment ,(2) Quality, (3)
Delivery precision and (4) Cost, see Figure 24 (Scania CV AB, 2010).
Figure 24 The priorities of Scania R&D
To measure the performance, Key Performance Indices, KPI:S, are generated for each one of them.
The result of the KPI:S is a basis for the improvement work. In an abnormal situation, compromises
among the priorities might be necessary and the order between the alternatives can help to act.
However, it is only in abnormal situation the decision principles are compared and eliminated with
this specific order.
3.4.2.4 Success Factors of Scania R&D
The R&D Factory house is surrounded by the three success factors of R&D; Leadership, Competence
and Creativity, see Figure 25.
Figure 25 Success factors of R&D
Leadership will ensure a positive interaction among managers, groups and co-workers. A manager at
Scania has the role as a coach; it should help the team to collaborate to outperform the individual. It
is also crucial for the manager to help the co-worker to grow as an individual. The leadership at
Scania has a focus on learning by doing (Scania CV AB, 2010). It is important to sustain a culture
where it is allowed to make mistakes. However, when a mistake occur some kind of learning has to
take place to ensure that it does not recur.
43
Competence is gained by both education and experience, but also by the individuals’ skills to conduct
its own work. The managers have to get the understanding for the future’s demand of competence
and ensure that Scania has core competence in strategic important areas.
Creativity is about thinking outside the box in new ways. Innovation at Scania is related to converting
a creative idea into knowledge that can create value for the end customer (Scania CV AB, 2011).
Therefore, creativity is a prerequisite for innovation. Creativity and innovation are not explicit for the
pre-study phase; they should be present in all steps of the product development process.
3.4.3 GEO
GEO at Scania is an abbreviation for GEOmetry assurance (Brantefors & Gadman, 2007). It is a
method used to ensure that each item has a feasible position to its surrounding components in each
truck variant, see Figure 26. The software is explicit developed for Scania. By using GEO it is possible
to explore deviations and interference issues in an early stage of the product development process
(Wernsten & Hanna, 2012).
Figure 26 Examples of GEO assurance for a battery box, the upper picture show the battery box itself, the
middle row pictures show three different possible surroundings of the box and the lower picture give an
illustration of the box when all possible surroundings are put together.
From a more practical point of view GEO links two different systems together; CATIA V5/Enovia LCA
and SPECTRA. The CAD model and its reference system, referred to as Part Axis at Scania, is collected
from Enovia while the items’ geometric position, GP, is collected from the system SPECTRA
(Brantefors & Gadman, 2007). It is the design engineer who updates the GEO archive by connecting
its 3D-model to its ID-number and GP (Wernsten & Hanna, 2012). The idea is that the design
engineer should do the first GEO update quite early in the design process. It might be just a quadrant
box that informs others that something is going to be in this area in the future.
44
GEO should be updated continuously; it should be an iterative process and should be done much
more frequently than other synchronization gateways in the projects (Brantefors & Gadman, 2007).
Today, there are no possibilities to see the maturity of an item in GEO. It is neither possible to see
older generations of an item as long as the item number is consistent.
GEO information is shared among several different functions at Scania, see Figure 27. Since it is a lot
of individuals who are using GEO in their daily work, it is of great importance that the version of the
item in GEO is up to date.
Figure 27 Information flow of the GEO archive, adapted from Brantefors and Gadman (2007)
45
46
4 RESULTS FROM CASE STUDIES
This chapter comprises the case study of this project. It has two main sections. The first one is based on
qualitative interviews at Scania and aims to describe the current state of the simulation usage at the product
development department. The second section of has a benchmarking focus and describes how three other
companies within the automotive industry have implemented simulation driven product development in their
product development processes.
4.1 Scania
Employees with different functions at Scania R&D Department have taken part in a qualitative study
in order to investigate the status of using simulation tools and simulation driven design at the
company. The study includes 25 Design engineers interviews, the questions is available in Appendix C
and a list of the interviewees in Appendix D. Six simulation groups have taken part and the simulation
questions is available in Appendix E and a list of the interviewees in Appendix F. Finally, three testing
groups and 20 other individuals who by taken part in interviews contributed to the understanding of
the current state, a list of the testing engineers and other individuals that has participated in the
study is available in Appendix G.
4.1.1 Interaction between Simulation and Design Groups

The design engineer at Scania has the full responsibility of its components. Therefore, it is usually up
to the design engineers to take a decision if simulations with help of the FEA or CFD engineers should
be performed or not. The simulation departments work as a support function and can sometimes
have responsibility for properties, but mostly they only give advice about strength and dynamics.
4.1.1.1 Simulation and Analysis Assignment Processes
The contact between a design engineer and the simulation department usually starts with a dialog
about what should be achieved. Although, to start a simulation job, a testing request, known as PA at
Scania, is written. The PA is used to keep track of all the jobs made by the simulation department. It
contains information about involved components, what questions or areas to investigate, which
project it belongs to etc. Except for the PA, there is no common method or process used at Scania
within the simulation departments. Some departments has made their own process, some wish to
have one, but do not have the resources to create one and some groups are so small that they have
not had the need for it yet. At the dynamics and strength analysis group for chassis, a process has
been made, see Figure 28.
47
CHAPTER 4 – RESULTS FROM CASE STUDIES
Figure 28 Work process at dynamics and strength analysis for chassis
The process described in Figure 28 comprises four different phases or sub processes; Initiate,
Execute, Report and Follow up. Usually the simulation assignment is made by iterations between the
design engineer and the simulation engineer. The iterations take place between the execution and
report phase. As a rule of thumb three iterations has to be done before the design fulfills the
simulation targets or requirements. Due to the iterations, one single PA can comprise more than one
loop of calculations. However, if it is a long time gap between the iterations it is preferred that a new
PA is used instead and the old one is closed.
When simulations are introduced in the PD-process is very individual dependent. Since it is a lack of
description about simulation and analysis in the design engineer’s check list, it is up to the design
engineer to involve simulation in the project. Therefore, when simulations are used in a project vary
from being introduced in the concept phase to not being included until just some weeks before SOP
when design usually s locked for changes. Traditionally, simulation has entered the projects too late,
when the components already has been broken in the physical testing or when the drawing has been
locked and only small changes are allowed. In projects with a tight time schedule simulation can
sometimes be made parallel with ordering physical prototypes which leads to that the simulation
results cannot be used to redesign before physical testing.
48
A simulation assignment can vary in time from one-two weeks to several months. The average
standard time for a fairly known component with a familiar simulation method takes about two
weeks. This time includes the start up meeting, modeling, calculation and analysis. The computation
processed by the computer can take from 20 min for a fairly easy FEA up to three days for a full
vehicle simulation. This time used to be longer and has been shortened due to increased computer
performance. Though, the simulation software has also become more advanced and therefore has
the calculation time not decreased as much the last few years compared to the increased capacity in
clusters and computers. If there is a possibility to do the calculation more detailed still within a
reasonable waiting time this is usually done. Most of the engineers working with simulation and
analysis prefer a more detailed calculation instead of reduced calculation time. FEA calculation has
reached the magic limit were they are calculated over the night and the result is shown for the
simulation engineer the next morning. As long as the result is shown the next morning, it would not
matter if the calculation takes two hours or eight hours and therefore the longer alternative is often
chosen in order to make the calculation more detailed.
In the end of every ordered simulation job from a design engineer a report is written. The report is
always connected to the PA number, contains the result and usually the method so information from
the simulation should be able to reuse. Some groups have started to write A3-reports with graphs
representing, for example, temperature distribution and fluid changes. This type of report is, in
comparison to a standard report, easier to overview and gives a faster understanding.
4.1.1.2 Simulations impact in the Product Development Process
The simulation result is almost always delivered before the report is written through a meeting
between the simulation engineer and the design engineer verbally so the design engineer does not
have to wait for the report to be written and approved to get the result. Every design engineer
interviewed is located close to their responsible simulation group which makes it easy to have a
regular verbal contact during the simulation assignment. The geographical position has a great
impact on the communication between the design group and the simulation group according to all
the interviewed engineers.
Simulation and analysis has taken a greater part in product development and has been involved
earlier in the projects since the design engineers has started to rely more on the simulation results
than before. Earlier, most simulations and calculations were made after the physical testing and their
purpose was to explain why the components broke or had cracks from the durability tests. It was
usually a very big hurry with these simulation assignments, also known as “fire fighting” assignments,
since the components were supposed to make it through the physical testing and the time to start of
production was not far away. The design groups with experience from these firefighting assignments
has gained increased trust and knowledge about the work of the simulation groups and therefore
started the testing for the upcoming projects with virtual testing and simulation instead for going
directly to physical testing. The case study has shown that groups with a history containing many late
rework projects and problems from physical testing has a bigger trust and reliability in the work and
results from the simulation groups.
49
4.1.2 Interaction between Simulation and Physical Testing Groups

According to the interviews in the study, the main purpose of simulation and analysis and physical
testing is equivalent. Both fields of expertise are striving to answer the same questions about the
product and its functionality. Though, their methodologies to come up with the answer differ. At
Scania, the two areas have a common history; simulation and analysis started to grow as a sub-
function to the physical testing groups. Therefore, simulation and analysis used to be done in the
same workgroups as the physical testing. Nowadays, the simulation and analysis groups tend to be
separated both geographically and organizationally from physical testing. Simulation and analysis is
treated as an own field with separate workgroups. Geographically they are often located more
closely to the design engineers than the testing engineers. Several of the engineers involved in the
study, both testing and simulation engineers, put attention to that they would appreciated if they
were located more closely and tighter connected to each other.
The simulation and physical testing groups highlight that they are complementary functions to each
other. Simulation and analysis is dependent of physical testing to validate their models. If the result
from physical testing and virtual testing mismatch, the fault is often assumed to be related to the
simulation model. Therefore, the results from physical testing can be used in order to do more
accurate models. The simulation engineer does not initiate a physical testing by its own to develop
methods and models. The design engineers tend to do both simulations and physical testing on the
same components which make it possible to get feedback from physical testing anyway. However,
how the information flows between the groups differ a lot. Some simulation groups have integrated
generating feedback within their groups’ process map and can by that ensure that they always get
feedback to its models and methods. Other simulation groups do not have a clear process description
and have not this opportunity. Instead they trust that the design engineers give them access to
feedback from the tests. Due to the common history of the two areas, it is a lot of personal contacts
between the groups which often is used to get feedback; the simulation engineer calls the testing
engineer in order to get feedback from the tests.
When the design engineer sends a concept to simulation it only gets the information that is asked
for. The output from the simulation is dependent on the questions the analysis and simulation
engineer wants to get answer to. The testing gives an impression of the entire picture since more
loads are present. Test engineers involved in the study highlights that testing should be called
measuring, since it is all about measurements. To measure a load or a stress, the sensors have to be
placed at right positions and the tests will actually only give feedback of what the customer asks for.
If the component breaks it will of course be detected without a sensor, but as long it does not break
it is impossible to get full control of what happens within the entire component. Physical testing is
dependent on simulation and analysis to be able to reach an accurate position of the sensors. Both
test and simulation engineers have recognized a trend of a lot of complex designs for basic
components such as brackets. The complex design makes it difficult to both do simulations and
testing in an accurate way and the complex geometry for the brackets probably depends on
increased modularization according to the testing engineers.
Sometimes is virtual and physical testing present in a project simultaneously and in parallel. It may
cause problem to get use from the result of simulations before the next generation of physical
testing appear, see Figure 29.
50
Figure 29 Example of simulation and testing simultaneously
Since the prototypes for the first run of physical testing is ordered before the result from first
simulation run is available, the knowledge gained from the simulation is not used in the first physical
test, see Figure 29. Interviewed testing engineers emphasize that they have several examples there
the result from physical test 1 is not available before the second run take place. This is a scenario the
groups would like to eliminate. In this case knowledge from the first simulation run is used for a
second simulation run. The final verifying test never occur in parallel with any other testing, neither
virtual nor physical testing.
4.1.3 The Role of the Design Engineer

The design engineer at Scania has several different kind of work tasks included in its daily work. A lot
of tasks are of course design related, but it also comprises administrative activities and to some
extent a role of a coordinator. The broad responsibilities and spectra of activities included in the role
of the design engineer sometimes make it difficult understand what the design engineer actually are
supposed to do in their profession.
4.1.3.1 The Daily Work of the Design Engineer
The work tasks and time share between them are quite individual. It depends on both the role in a
project the design engineer has and how it wants to work. Some design engineers are interested in
simulation and analysis and tries to do as much as possible by themselves while others feel
inconvenient when getting in touch with analysis and simulation tools. Some design engineers do all
CAD work by themselves, while some design groups have consultants who handle the work in CATIA.
However, almost all design engineers involved in the study mention their component responsibility
first when they are asked about work tasks. Most of the design engineer has responsibility for one or
several items. Explicit what kind of activities and tasks involved in the component responsibility
depends on how mature the component is. Some design engineers are involved in projects with new
concept generation and tries to come up with new concept to solve problems while other engineers
responsible for mature components conduct and fine tune the current component.
The cross functional work, as contact and interaction with other people involved in the project, is
another recurring task for the design engineer. The majority of the design engineers spend a lot of
time in meetings or in discussions with, for instance, production, purchasing or subcontractors. Some
of the interviewed engineers highlights that they spend a lot of time distributing and searching for
51
information regarding their specific component. This type of activity has often the same purpose as
the meeting, i.e. gather and distribute information. Help other engineers or asking for help are other
frequent activities mentioned as a part of the daily work.
The documentation is included in the daily work for a design engineer. Almost all engineers
mentioned the tasks connected to create and maintain Engineering Change Orders, ECO:s, related to
their component as a significant part of their daily work. The attitude to documentation diverge;
some engineers emphasizes it as something good and a part of their learning process while other
engineers describe it as time consuming administrative tasks and doubt the value of it.
Some design engineers also mention the continuous improvement work and enhancement of
methods as a part of their expected daily work. Although, they say that they should do it, no one has
actually put attention to it as a frequent part of the daily work.
4.1.3.2 Design related Time
As a quantitative part of the qualitative study each design engineer was asked to estimate what
share of their time was related to design tasks. Design tasks are not only related to CAD. The
participants in the study were asked to bring all activities related to add value to its component into
consideration when doing the assumption. The result gives a mean value of 35 % and a median of 30
% time spent on design related tasks. The entire distribution is visible in Figure 30.
Design related Work Tasks

Estimated share of work time
1
0,8
0,6
0,4
0,2
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Individual answer from each interviewee
Figure 30 Design tasks’ share of total time at office
Several of the design engineers emphasize that the share of design related work tasks tend to
decrease proportionally to the employment time at the company. This hypothesis is supported by
the result in the study, where senior engineers have other complementary areas of responsibility and
also tend to get interrupted more frequently by questions than an engineer in an early phase of its
profession.
4.1.3.3 Knowledge Developing Activities
What kind of activities enhancing the knowledge of the design engineers seems to be quite
individual, see Table 7. Some design engineers increase its knowledge when they can work in CATIA
undisturbed for a longer period. Others gain knowledge by cross functional discussions and feels that
working in CATIA is more of a repetitive kind of work task.
52
Table 7 Knowledge developing activities mentioned by design engineers, all engineers were allowed to
emphasize unlimited number of activities. Therefore, the sum of the sample share is more than 100 %
Knowledge developing activities Share of sample

Feedback from testing and simulation 32%
Cross functional discussions 32%
Working in CAD by my own 32%
Own simulation and analysis 20%
Discussions with colleagues 20%
Collecting information/Reading reports 12%
Benchmarking 8%
Commitment to the work task 4%
Everything that is a part of my work
(Including administration) 4%
According to the discussions with design engineers and the statistics presented in Table 7 what
knowledge developing activities are seems to depend on two factors. The first one is what kind of
work task the certain design engineer has and what kind of obstacles it faces in its daily work. The
second is how it defines knowledge developing. The definition of knowledge was dependent upon
the answering design engineer which may have influenced the deviation in answers.
4.1.3.4 Secure “Right from me”
Right from me is one out of four principle of R&D Factory at Scania, see section 3.4.2.2. The purpose
of the principle is to ensure that an activity is executed and delivered to next internal customer in the
right way at the first time. The design engineer uses different methods to fulfill the right from me
principle. The most trivial one is to follow the designers’ checklist which includes several steps and
activities that should be taken into consideration in the different design phases. However, several
design engineers find the level of detail of the checklist to be unleveled. Some steps are described in
detail while others are widely opened for individual interpretation. One example is the step
“brainstorming” which can be done in unlimited different kind of ways. On component level, design
engineers uses designer guidelines which is a document including learning from previous work done
regarding the component. The guidelines are used more frequently among recently employed
engineers in comparison to senior engineers. When an engineer has looked at the guideline a couple
of times it keeps the most important in mind. Another tool to deliver right from me is to follow
standards such as PA when sending a request to physical testing or simulation and analysis.
When mention right for me, the majority of the design engineers relate to production and how the
quality of the component they are responsible for is secured before it is launched to market. Physical
testing is the primary way of secure the quality of the component before a delivery. Several, but not
as many as uses testing, also uses simulation and analysis to ensure that the component fulfills the
targets. Some of the engineers that uses GAS to do some FEM simulation and analysis by themselves
emphasize that GAS give them an understanding for the properties of their component and by using
53
GAS in collaboration with a simulation and analysis engineer they can secure that they deliver right
from me.
Approximately 15-20 % of the design engineers that have taken part in the study highlights that they
do not follow any standard or method to secure right from me. Some base their concepts on intuition
and experience while others emphasize that they actually do not know that they deliver right from
me, instead they try to do as good guess as possible.
4.1.3.5 A Design Engineers Consciousness of Economics
The parameter almost all of the asked design engineers start to think about when talking cost is the
manufacturing cost for their component. The design engineer tries to make a smart construction and
avoid expensive and unnecessary tools in the manufacturing process and find as cheap raw material
as possible. However, some of the engineers involved in the study emphasize that the engineers
should put more attention to weight optimization when designing; the truck weight influence both
fuel economy and load capability for the customer, i.e. it affect the customer value. The weight also
affects the need of raw material and therefore the component cost at the market.
When focus on development process cost for the components, the awareness is much lower. During
the interviews, the development process cost has been defined as the costs for developing the
component. It involves factors such as expenses for prototypes, physical testing and internal
resource consumption, when for instance order a simulation work from another person. The majority
of the design engineers tend to ensure that the right number of prototypes are ordered, but less
than ten percent emphasize they actively tries to avoid, for instance, unnecessary physical testing
due to a believe that physical testing is expensive. One of the design engineers says that it believes
the way of reducing cost is to do as much as possible by itself, for instance to do FEM calculations in
CATIA. According to this engineer the cost heavily increases as soon as a hand over to another person
takes place.
The cost for prototypes is displayed to some design engineers, but no one involved in the study can
connect the cost of prototypes to the total project cost. Several design engineers says that they have
a lack of feeling for cost; a prototype might cost 100 000 SEK, but they cannot estimate if it is a small
or large share of the project budget. The majority of engineers answering the question come up with
the conclusion that they do not have any connection to cost and therefore a lack of consciousness.
Another statement among the answers is the lack of incentives for reducing development process
cost.
No one could give an accurate number for the cost of delay of a project to start of production. They
tried to think of consequences, but could not give a monetary term. However, some tried to guess a
value, but they clarified that they did not actually had a clue.
54
4.1.4 Simulation and Analysis executed by the Design Engineer

Of the interviewed design engineers almost everyone has been in contact with the Finite Element
Analysis, FEA, application GAS in CATIA. Most of them got familiar with it recently and has still not
used it as much in their daily work. Some design engineers believe that how frequent GAS is used
depends on the background of their group manager; if the group manager has a background as a
simulation engineer, GAS is encouraged and used more frequently.
4.1.4.1 The Experiences of Simulations executed by the Design Engineer
Many design engineers find it difficult to get started with the GAS application. They experience that
the entry level is pretty high, especially if you have little experience from FEA software and solid
mechanics from earlier. Therefore, they need help and support from their simulation group,
especially in the beginning regarding loads, boundary conditions and how to build a FE-model. They
also feel a need for help to evaluate the results and decisions how to proceed. During the evaluation
of the GAS calculation, the simulation engineers can also take the decision if they need to do a more
detailed simulation in their group or if the GAS-simulation is good enough to continue to the next
step, i.e. physical testing. Design groups that do not get support from their simulation group
regarding GAS, does not use GAS as frequent as design groups that have a supporting simulation
group who can help them.
The design and simulation groups are currently often located next to each other, which almost every
design engineer mentioned as helpful. It is easier for the design engineer to have a dialog with the
simulation group when they are familiar with each other. They can have the discussion about the
results from either the GAS simulation or a regular simulation at someone’s screen instantly instead
of waiting for a meeting setup or until the report is finished to get feedback.
The GAS work made today is not as well documented as regular simulation jobs. Some design
engineers do some print screens and others automatically generate a report inside CATIA and GAS.
Though, the automatically generated report does not contain any valuable information according to
the interviewed simulation engineers and design engineers with more GAS experience. Most of them
also mentioned that if GAS becomes more common and is used as decision making support a report
is needed for every GAS work made.
4.1.4.2 The Potential of Simulations executed by a Design Engineer
Both the design engineers and the simulation engineers that have experience from working with
GAS, emphasize that the GAS application should be used to see differences in strength when
comparing two designs and not to find any absolute values . Therefore, it can especially be used for
components that have field quality problems today and needs to be more robust, but without
increasing weight.
According to the interviews GAS is very helpful when it comes to design new or improve already
existing brackets. Brackets have not been seen as components that need to be simulated compared
to the silencer, fuel tanks or electronics that the brackets usually hold. Therefore, the feedback for
the design has not arrived until after the physical testing. GAS has therefore been useful to design
the brackets without guessing when it comes to ordering prototypes for physical testing. GAS can
also function as an effective interface between the design engineers and the FEA engineers,
according design engineers taking part in the study. Since the design engineers get more familiar
with analyzing FEA color grids and red spots in their calculation model, they also feel like they
understand the discussions better regarding more complex FEA work made by the simulation
55
engineers. The simulation groups do also experience more quality in the design when taking over a
model that has been calculated in GAS. They do not need to do as much iteration compared to if GAS
has not been used.
GAS is more popular among design engineer with little or none experience and younger employees.
Since a GAS simulation goes a lot faster than a regular FEA software simulation the design engineer
can get a feeling about the strength and robustness through more iteration many weeks earlier than
by doing a regular simulation. By using GAS, the quality of the components get enhanced very quickly
since the design engineer knows more about the design and its strengths and weaknesses than it did
without using GAS. Many of the design engineers said that they often had no clue how good and
robust a design actually was and they often draw new geometries based on guesses combined with
earlier versions and reports from the field. They did not really get a feedback on their design until
they started the physical testing. If they had bad luck and their guess did not make it, the rework was
usually very expensive, time demanding and stressful since rework is not a part of the project plan.
Another issue has been to gain knowledge from simulations.
Often the simulation department delivers the results from the simulations as a yes or no. By using
GAS in the early design phase, the design engineer can both reduce the workload of the simulation
groups and gain more knowledge about how different design parameters affect the robustness.
4.1.5 Synchronization and Geometry Assurance at Scania R&D

Today, most simulations are made on components or smaller sub-systems. The interviewed
engineers mention that full vehicle simulations are rare since they do not have any virtual prototype
vehicles to simulate and analyze. The current Product Data Management, PDM, system containing
the CAD models and their GEO-position does only comprise the latest updated model and therefore
they cannot go back and use an older model to build a virtual vehicle. It is also difficult for the
simulation engineer to know if the model in the GEO archive is the current valid model right now or if
the design engineer is working on a new model that has not been updated. The simulation engineers
that were interviewed feel a need for some type of synchronization for virtual geometries so they
knew when they could go in the GEO archive and pick out the models. Both the design engineers and
the simulation groups feel a lack of expressing maturity levels in the models so other users could see
if the model in the concept phase, under development or locked due to tool order.
4.2 Benchmarking of Simulation Driven Product Development

To get an enhanced understanding for SDPD in the automotive industry in general a benchmarking
study at three other companies are made. The first company that is studied is former SAAB
Automotive AB. Due to the bankruptcy in December 2011 the case is based on a story from former
employees. The case describes Saab’s journey of reducing the amount of prototyping heavily in favor
of simulations. Next case is about Jaguar Land Rover, JLR, a classic British car brand, nowadays a part
of Tata Motors. According to the simulation software provider Dassault Systems JLR is in the cutting
edge of using simulations in its product development process. The third and final case is about AB
Volvo with focus on how its four truck brands use simulations in a common global development
process. Volvo is analogues to Scania providing heavy duty trucks and has Sweden as base for its
headquarter. Therefore, a benchmark study of how Volvo uses SDPD is of interest for the study.
56
4.2.1 Cost Rationalization by using Simulations at Saab Automobile AB

Saab Automobile was under hard pressure from General Motors, GM, to reduce cost by reducing the
number of physical prototypes and testing. Therefore, virtual development without physical
prototypes was implemented. Both the Saab 9-5 MY10 and the new phoenix platform for Saab 9-3
were developed without usage of classic physical prototypes. However, the mule vehicles were used
to test new combinations with existing components from the platform shared between several GM
brands, for instance the Saab 9-5 and Opel Insignia, and one production validation series were made
with components adapted for mass production. Regarding the new 9-3, the validating production
series never took place since Saab Automobile went bankrupted in December 2011. Saab Automobile
used a platform strategy together with the rest of the GM Group which helped them with the virtual
development since many of the parts used in the 9-5 had already been tested physically for other
vehicles. The strategy was to remove the alpha-, beta- and gamma-prototype series, integration
series where the prototypes are made from scratch, and replace them with virtual testing, see Figure
31.
Figure 31 New test plan for the development of the new Saab 9-5.
The reactions were strong, especially among employees within the physical testing and with property
responsibility. According to them, it was impossible to replace all the physical testing since it was a
too big risk and it would lower the quality. Therefore, every property responsible had to go over their
tests and prototypes to replace the physical testing with virtual methods. Every property responsible
still had to have physical prototypes, but since the Saab management was under hard pressure from
GM to reduce cost and lead time they did not accept any prototypes they had to develop new
methods. The physical tests were either replaced with a simulation method or with a test result from
the platform. Every test that did not have a virtual test method got a Road to Lab to Math, RLM,
project. These projects were supposed to find a way to move the physical road tests in to the lab, still
with physical prototypes. Once the lab tests were verified, the next step was to remove the physical
prototypes and replace the lab test with a virtual test. Sometimes it was even possible to go directly
from road testing to a simulation method. In several areas, a simulation tool was missing, so Saab
had to involve their software providers in order to develop new applications.
The big challenges with the RLM projects were to translate different complex feelings, for example
the feeling of closing a car door to measurable parameters from the simulation. In order to be able to
translate the attributes based on feelings and driving performance you need to have experienced
employees with knowledge about what to looking for while testing prototype cars (Sjödin, 2010).
Therefore, the biggest challenge area with virtual development was vehicle dynamics, since it is hard
to translate the feelings from driving a vehicle in to measurable numbers. It was important to break
57
down big goals to measurable parameters, for example five stars in EURO-NCAP had to be translated
to acceleration levels for different components and “a comfortable sound level” has to be broken
down to Eigen mode frequencies (Sjödin, 2010).
Moving over from physical to virtual testing also required an organizational change since they
needed more simulation engineers and less lab test engineers. Many lab engineers were a big help in
the RLM projects and continued working with the tools they helped to develop. Saab also needed a
more centralized CAE group. Their earlier organization had the CAE- engineers spread out in the
different areas which gave problems with non-synced CAD geometries and simulations were made
on a model at the same time another group made changes on the same model. Saab kept the CAE
engineers decentralized, see Figure 32, but they reported to a centralized CAE group which took care
of the models so everyone could be aware of any changes that was made.
Figure 32 New CAE organization (Sjödin, 2010)
The physical tests, mule-, alpha-, beta-, gamma- and the validation test series, had become the spine
in the product development process at Saab automobile. All the other work, like simulation and
analysis, were just added to the test series process. The big problem with the physical test series was
that the results and changes from the alpha series were not tested in the beta series since the
prototypes for the beta series had to be ordered long before the alpha series was finished. Many
problems therefore remained through the beta series even though they were already known and
discovered in the alpha series. Since a prototype series takes six months it is very time demanding to
go through the testing without being able to include the change made from the earlier series. It is
also very expensive to test prototypes one more time that already have been tested and where the
problems already are discovered.
When moving over to virtual testing the series testing phase were minimized to 6-8 weeks instead of
6 months. To be sure that the tests always were made with the latest version of geometry, they used
synchronization time points there the CAD models geometry were frozen. This way, one simulation
group, for example noise and vibrations, were sure it was the right model they were using since no
other group, for example safety, were able to make any changes in the geometry. When the
simulation was done, they held a vehicle assessment meeting during 1-2 days where all the problems
and issues were brought up. After the meeting the simulation engineer worked together with the
design engineer to solve the problems until the next sync time point, see Figure 33.
58
Figure 33 Time plan between sync points (Sjödin, 2010)
This new strategy made the synchronization points and the vehicle assessments to the new spine in
the product development process. With less hardware prototypes resulted in less administrational
work and the design engineers therefore had more time to spend on the design instead of fill out
paper work for ordering prototypes.
Their next step with SDPD was to shorten the time between the sync points and Vehicle Assessment
meetings with help of a more automatic process for the mesh modeling, CAD assemblies and the
evaluation of the results, see Figure 34.
Figure 34 A more automatic process between sync points and vehicle assessment (Sjödin, 2010)
Another area that they wanted to look further into was to use optimization tools in the phase after
every vehicle assessment to shorten the development time even more.
The big advantages were the reduced overall cost for the development process, the shortened time-
to-market and the enhanced quality on the product. The Saab 9-5 received five stars in the EURO-
NCAP safety test and it was development with less than 50 % of the physicals tests compared to the
previous 9-3 model. At the same time the number of requirements and workloads were increased.
There were also less late issues to solve in the project during the validation series compared to
earlier projects. According to Sjödin, the most important factor for moving over to virtual driven
development is the have the right methods, an organization with skilled and brave employees and to
have a supporting process integrated with the simulation tools.
59
4.2.2 Towards Zero Prototyping at Jaguar Land Rover

Jaguar Land Rover, JLR, has had clear business goals due to increased market competition. Except for
just delivering premium vehicles with reduced CO² emissions, they always want to deliver a robust
design for all world markets with a reduced dependency on physical prototypes and a shorter
product development time (Boon, 2009). In order to reduce physical prototypes, JLR will use more
Computer Aided Engineering, CAE, and increase the confidence for CAE. In the mean time, their
increased CAE usage is supposed to generate fast design iterations which will lead to shorter product
development time. More detailed CAE is also supposed to help them get a more robust design (Boon,
2009).
Jaguar Land Rover, previously owned separately by Ford, is now going through an offensive system
platform exchange called “End to end product development with a single architecture”. Jaguar and
Land Rover were acquired by Tata Motors in 2008, they required JLR “Clone and Go” which meant
that all the data had to be copied over to their own systems and JLR was not allowed to work at the
same system platform as Ford anymore. The applications and systems used by Ford were both in-
house made and delivered from many different suppliers. This made it very expensive for JLR to just
maintain support for their old system platform and the applications from Ford. It was also far from an
ideal platform for JLR since they had very different sales volumes compared to Ford; 250 000
vehicles/year versus millions of vehicles/year.
JLR divided their system platform problem in to four pillars.
Diagnose Need for Change Vision Implement
•Current state •Cost to serve the •Zero prototyping •New system

analysis current platform development platform and
•Competitors •No transparency from 2013 migrated data in
Status in information •Information Q1 2013
sharing transparency
•Waste reduction
Figure 35 Four pillar of the ”End to end product development with a single architecture”
The competitors status in the diagnose pillar, see Figure 35, was made as a benchmarking matrix
where different areas as digital manufacturing, PLM, visualization etc. were on the horizontal plane
and knowledge level on a scale from one to five on the vertical plane. The goal for JLR was to be as
good as the best company in every area.
Together with this project, they created a 50-30-10 strategy where their goals are to have 50 % more
products, reduce time to market with 30 % and decrease the tool changes due to late design changes
with 10 %. Combining this strategy with the goal of being the best in all the different simulation areas
shows that they are clearly choosing a SDPD.
Sustainability has also been a driving force behind a simulation driven process. When developing the
new Jaguar XJ virtual testing performed 7000 crash tests and over a million miles of virtual driving
which gave them far more data and information then what physical testing would have (CIMA, 2009).
It also reduced emissions and cost within their product development process. Then introducing more
60
virtual methods instead of hardware prototypes JLR has had the sustainable benefits in focus as well
as the economical. It did reduce the cost for their prototype manufacturing with 50 % between 2002
and 2003 and a recent modular project used 93 % less hardware prototypes compared to the
previous one (Jaguar, 2012).
4.2.3 Simulation Development Loops at AB Volvo

This section is based on an interview at Volvo Group Trucks Technology. Volvo Group Trucks
Technology is a subsidiary of AB Volvo and supports the four AB Volvo truck brands Volvo Trucks,
Renault Trucks, Mack and UD with product development, planning and purchasing services.
AB Volvo used to have difficulties to perform simulations of the entire truck due to lack of
synchronization in product development projects. Everyone worked as fast as possible, but it did not
create any synchronization. Due to this, it was almost impossible to get sufficient data to perform
simulations with the entire truck as boundary boarder. The solution to the problem has become
development loops. It was launched in 2007, but since the time horizon to develop a new truck
generation is fairly long the development loop process has only been completed once.
One development loop comprises three different phases; (1) Definition, (2) Implementation and (3)
analysis, see Figure 36. The definition phase includes both definitions of what to evaluate during the
loop as well as evaluation of the results from the previous loop, this phase last for 4 weeks.
The implementation phase is the most time consuming, it takes approximately 20-25 weeks, and it is
in this phase the majority of the development takes place, each time this phase is performed for a
component its maturity will increase and a lot of simulations at component level are performed
within the implementation phase. In the analysis phase simulation on the full vehicle level takes
place, it is a gateway for the design engineers and all components have to fulfill a predefined level of
maturity. The analysis phase takes place in parallel of the definition phase of the next loop. A project
consists of several steps. The last project using the development loop process at Volvo had seven
loops, each one of them with a predefined purpose. The development loop project is performed by
eleven cross commodity groups with shared responsibilities, four groups focus on the cab and seven
on the chassis development. The simulations, that have the entire vehicle as system for the analysis,
are performed by simulation engineers with ordinary location at chassis simulation.
Figure 36 Illustration of Development Loops at AB Volvo
61
4.2.3.1 A Way of handling Maturity of Components

Volvo categorize its components that are included in a project into three different categories; (1)
Prioritized Critical Areas comprises development of new components or technology with large
importance for the success of the entire truck. (2) Critical Areas comprises development of new
component with less importance or complexity in comparison to the prioritized ones and the
category (3) ALL comprises all other components included in the project, where are often a lot of
“carry over” components from previous generation. All components are classified into these
categories by a cross functional network. Within each category the components have three levels of
maturity, LOM, which are defined as E, F and G, see Figure 37. These three steps are defined for all
CAD-modules. A CAD-module can for instance be the exhaust pipe from the engine to the muffler
including brackets and fasteners. Next module might be the mufflers itself inclusive brackets and
then can the tail pipe of the exhaust system be a third module.
Figure 37 Illustration of CAD-geometry’s level of maturity
A LOM-E gives the basis for geometry envelopes for each module to claim the space for new design
modules. It is usually a compromise between different modules to perform as good designs as
possible; almost all design engineers want to use more space than what is available. A module with
LOM-F has locked it outer interfaces and at this level some simulations regarding the complete
vehicle or sub-systems can be made. However, it is first when the modules have reached LOM-G
status the simulations regarding strength and dynamics can be made on whole vehicles. The
prioritized critical areas of the truck reach a higher level of maturity before the other categories. One
feasible focus of the first development loop might for instance be that all prioritized critical areas
fulfill LOM-E.
4.2.3.2 Physical versus Virtual Testing at AB Volvo
As mentioned earlier in the thesis, simulation and analysis is virtual testing often with a common
objective as physical testing. At Volvo the view of testing has changed. Nowadays, almost all
developing testing is made in a virtual reality. Simulation and analysis is performed on an item until it
fulfills all targets theoretically and then the physical tests are only used as validating tests to secure
that the simulation model was accurate. It is no time for surprises in the physical tests, though it
happens sometimes, it does not appear often enough to be a disturbing problem for Volvo. However,
62
it is the design engineer who is responsible for the quality of its component and if it feels secure that
the component fulfills the targets by previous experience it actually does not have to do any
simulations. In case of insecurity of the strength of the design, simulations are the way of evaluating
it. It is important with a tight collaboration between physical testing and virtual testing, measures
from the physical testing track are used to update simulation models and verify its accuracy. The
virtual tests and the validating physical test are usually consistent.
Volvo has a vision of zero prototyping, but today still it is only a vision. The focus is to find the
shortest route to the answer. Sometimes it would take ten times longer time to simulate a result
than ask the mechanical workshop for help with a simple prototype. In this case the physical
prototype should be favored. It is mainly in two areas the virtual tests do not reach sufficient
accuracy in sufficient time; how the wire harness behave when the cab is tilted and routing of wire
harness. These two areas still have a need of physical testing. On a larger perspective Volvo has a
potential of implementing simulation tools to a larger extent in the pre-production. A lot of assembly
preparation operations are still performed physically even though it is virtual tools available.
The work with development loops has had a great impact on the product development lead time.
When the project that has had this loops and simulation on the full vehicle level is compared to the
classic global development process the lead time has been decrease by 20-30 %. The development
loops is the new standard procedure at Volvo and it is actually founded on the basic principles from
the middle of the 20th century to front-load product development projects as much as possible.
63
64
5 ANALYSIS AND DISCUSSION
In this chapter, the findings from the theoretical framework and results from case studies are analyzed and
discussed. The analysis has to some extent a Scania perspective, but will be applicable generally as well.
5.1 Feasibility of combining Lean and Simulation Driven Product

Development
To investigate to what extent Simulations Driven Product Development, SDPD, can be integrated in
Lean product development an analysis have to be made in several levels. The Lean philosophy and
SDPD methodology has to be compared in several different layers. On an upper level the different
principles of Lean product development might be affected by SDPD. On a lower level it is more a
matter of how Lean methods, such as SBCE, are affected by an implementation.
5.1.1 The impact on the Lean Product Development Principles.

The impact on the thirteen principles of Lean product development introduced in section 3.2 Lean
Product Development will be analyzed from the perspective of the three categories process, people
and tools and technology.
An implementation of SDPD might affect principle 2: Front-load the product development process
while it is maximum design space to explore alternative solutions thoroughly. In comparison to
testing, simulation and analysis can be performed on an earlier stage since it does not demand
physical prototypes. This should facilitate the front-load principle. However, it is very dependent on
how simulation tools are used in the process. In order to support front-loaded product development
processes, simulation tools should enable to base decisions on facts. It can for instance be used in
the pre-development phase when concepts are going to be evaluated. Sometimes simulations is used
for firefighting and backtrack an unexpected failure of a component, this is not related to front-
loading at all. As long as simulation tools is used proactively to drive the product development
forward it will have a positive contribution to the second Lean principle introduced by Morgan and
Liker (2006).
An implementation of SDPD might affect two of the principles connected to the area people. The first
is Principle 7: Develop towering technical competence in all engineers. In cases there the SDPD
comprises simulations performed by the design engineer it might be seen as the design engineer
starts to broaden its competence by starting using simulation tools instead of increase competences
in CAD. To some extent simulations performed by design engineers might be a contradiction to
principle number seven. However, it is all a matter of what kind of competence the design engineer
role conduct. If the design engineer is supposed to have skills in strength and durability for its
component simple simulation might be useful to gain this knowledge and can be classified as a
towering instead of a broadening activity. Since strength properties often is an important target
regarding the design, it should be reasonable to combine principle 7 and simulation performed by
the design engineer. Other kinds of simulations does not have an impact on this principle, but if a
simulation driven approach is chosen it will be necessary to have engineers who are experts and
towering within the field.
65
CHAPTER 5 – ANALYSIS AND DISCUSSION
SDPD might also affect Principle 9: Build in learning and continuous improvement. In the case where
the design engineer performs simulations, learning connected to strength theory is going to be built
in the system. In order to support continuous learning simulation results have to be well
documented. Once again the impact of an SDPD implementation is a matter of how the
implementation is done rather than if it is implemented or not.
When studying the principles associated to tools and technology it is more related to how to
implement SDPD in a Lean environment rather than a matter of if Lean production and SDPD support
each other or not. To be supported by these principles, implementation of new simulation tools has
to be performed in alignment with the people and process. It is important that simulation tools are
used to support the process and people and not vice versa. Therefore, it should be reasonable to be
a little bit conservative when implementing new technology and tools in order to assure that the
people and process always are in focus.
5.1.2 The impact on Waste in the Product Development Process

In the theoretical framework waste has been defined in several different ways, see 3.2.2 Definitions
of Waste. The categories introduced by Ward (2007) is developed for the product development
environment explicit in contrast to categories introduced by Morgan and Liker (2006) that is more an
adaptation of well known waste categorizations from Lean production. Scania R&D has its own
categorization of waste even though it is quite similar to the categories defined by Morgan and Liker.
5.1.2.1 Scatter, Hand-Off and Wishful Thinking
An implementation of SDPD will probably affect the amount of scatter waste in the product
development process. However, the impact is ambiguous; it is a potential of decreasing the amount
of waste, but it is also o risk that a SDPD implementation actually increasing the amount of waste in
the process. If design engineers starts to use for instance simple FEA tools to do some simulations by
themselves it is likely that their understanding for the simulation engineers environment and
preconditions will increase. Another positive aspect is that to be successful with simulations the
design engineers probably need to perform its simulation with support by a simulation engineer;
therefore, the frequency of the daily contact between the different workgroups should increase.
These two aspects correspond to decrease the part of scatter waste related to communication
barriers.
When implementing SDPD the available research suggests a kind of method standardization. The
result should be analogues and independent on the individual who perform a task. This approach
initiates a risk of forcing engineers to use tools that is not sufficient for its certain situation; the kind
of scatter Ward (2007) define as waste caused by poor tools. It is only engineers who get usage for
the result of the simulation that should use SDPD and therefore it has to be done some examination
of which circumstances the method should be used or not. If a too detailed SDPD methodology is
implemented all over the department it will probably cause this kind of waste.
Waste caused by Hand-off will probably be affected by a SDPD approach. The first root cause of
Hand-off Ward derives is useless information, i.e. generated information that does not add value. If a
design engineer executes simulations by itself it will get only the answers it asks for and it can be
seen as a method to secure that useless information is avoided. However, it is a risk that the accuracy
of the calculation or simulation performed by the design engineer is low and actually does not add
any value to the quality of the design. In this case SDPD actually increase the amount of waste in the
process. A trend during interviews with simulation engineers was that they prefer to come up with a
66
result that is as close to the reality as possible. However, sometimes it is not a need of high accuracy
of the computation or simulation. In cases were a simulation is more accurate than needed waste
has actually been added to the process due to that the simulations have used more resources, both
man hours and hardware, than actually was needed for the computation.
Ward classifies waiting as the second root cause of Hand-off waste. The waiting is twofold; it is
related to a work-in-progress project have to wait between value-adding activities as well as
individuals hand over its responsibility to another division and waiting for input to continue their
work in the project. The possibility of decreased communication barriers between simulation and
design engineers due to enhanced understanding for simulation among design engineers will
probably reduce the waste of individuals waiting. It will increase the cross functional work which
Ward suggest as a solution to the waiting issue. For the other part of waiting simple simulations
made by the design engineer reduce the number of transfers between different individuals. As soon
as an order is placed to a simulation group the object has to queue for a free simulation slot. Since
the number of iterations between departments likely decrease SDPD has a possibility of decreasing
the waste caused by waiting.
The final category of waste according to Ward (2007), wishful thinking, is probably not as affected of
an SDPD approach as the other categories. For instance the behavior of testing to specification is
independent of if the test is executed virtually or physically. However, it is essential to put attention
to the importance of deliver more knowledge than just a binary answer if the design fulfill the targets
or not. Ward suggests an approach where tests should provide the design engineers with knowledge
and possible solutions. It is a statement to keep in mind when implementing SDPD and developing
methods and processes in order to gain knowledge. The second root cause of wishful thinking is
related to how the knowledge gained during tests is sustained. If results from simulation is not
documented and reused when a comparable simulation takes place next time it will drive waste
related to discarded knowledge. Therefore the development process should include demand of
sufficient documentation and to sustain the knowledge from a development cycle or an individual to
another.
During the introduction of the thesis rework is assumed to be one of the most common wastes at
Scania’s R&D department. Even though rework is not stated as an example of waste by Scania (2010)
it is mentioned during interviews with employees as a common waste. However, rework in itself is
not obvious to assign as pure waste since it at the current state sometimes seems to be necessary in
order to deliver demanded quality. Therefore, rework in itself is maybe not waste, but rework is a
clear sign of other wastes present in the product development process. Therefore an implementation
of SDPD should reduce the number of rework cycles.
Changes towards SDPD, from a more traditional development procedure, will probably have an
impact on all Ward’s waste categories. However, if the effect is going to be positive or negative is
probably closely related to how SDPD is implemented to the product development process.
67
5.1.2.2 R&D Factory’s Definition of Waste

In comparison to the categories Ward prefer for classifying waste, the examples of waste described
by Scania (2010), see 3.2.2.3 Scania R&D Factory’s Definition of Waste, is not affected by a SDPD to
the same extent, although, it might have an impact on some of the categories. Doing more than
necessary might be increased in the same way as it might cause waste caused by hand-off in the
previous section. However, if some simulations and calculations are suitable to be performed by a
design engineer instead of a simulation engineer it is also a possibility of avoiding more accurate
simulations than actually needed. This is also related to the example Not using everyone’s
competence. When a simulation engineer perform simulation that actually is suitable to be
performed in an integrated simulation module to the CAD software by the design engineer it should
be waste of the simulation engineers competence, due to overqualified skills. Instead the simulation
engineer could have spent its time on more complex tasks, which for the moment is idling in a queue.
However, the phenomenon described in this section is more a matter of how simulation is used in
the PD process and of who actually performs the simulation than if the product development process
is simulation driven or not.
5.1.3 Flow and Cadence

As mentioned in the theory framework chapter the flow is in focus in a Lean enterprise. The target is
to generate a seamless flow where the amount of waiting and other waste is minimized in
contradiction to the traditional approach where focus is on maximize resource utilization. To sustain
the flow the presence of an obvious cadence and synchronization will be necessary. From a SDPD
perspective several individuals independently emphasize the need of cadence and synchronization to
be able to do simulations that comprises several different sub-systems that affect each other.
Therefore, SDPD should be more easily implemented at a company that already has implemented
Lean production and has initiated some kind of flow in its product development process than in a
company that has not. Still it seems to be common that it is either the start of production or planned
physical tests that steering the product development process. The projects are planned with these
deadlines in mind and it is dependent on available time to what extent other activities, such as
simulations, are included to the plan or not.
In a SDPD approach simulations should be a key to both create the flow and identify the cadence and
therefore, simulations probably has to get a more central role than it has at a company that currently
has physical testing as primary gateways during the process. Several important simulations in a SDPD
environment will take place on a lower level, for a component or subsystem, than the large gateways
in a flow driven by physical tests or start of production. Therefore a cadence on project level which
Ward (2007) argue for should be suitable for a SDPD approach. A project individual cadence would
level the work load on different functions since all activities do not occur simultaneously. However,
simulation with large system boarders such as full vehicle simulations probably has to occur
synchronized and with a common cadence for the entire department.
68
5.1.4 Set-Based Concurrent Engineering

Set-Based Concurrent Engineering, SBCE, introduced in section 3.2.3, is on a more detailed level than
the philosophy and basic principles discussed in previous sections of this chapter. From a holistic
view the core of SBCE is two principles; base decisions on facts and eliminating concepts stepwise.
One common opposition to SBCE has during interviews been that it is to resource consuming to work
with several concepts too long into the project. If each decision has to be based on fact all concepts
has to be evaluated to some extent. In a physical test oriented environment it will be very costly and
time consuming to do prototypes and actually test concepts to eliminate them.
With a SDPD approach it is actually possible to do calculations and simulation at an early stage of a
concept and therefore base concept elimination on fact instead of intuition. If the early decisions in
the product development process are based on fact instead of intuition or best guess it is likely that
the number of rework cycles due to insufficient design will decrease. In cases, such as the Saab case
described in section 4.2.1, where SDPD is implemented to an extent that makes it possible to
eliminate all physical tests except verifying tests parallel concepts can be used all the way. Therefore,
SDPD and SBCE is consistent, actually if SBCE are going to be used in an environment as complex as
the automotive industry, SDPD seems like a precondition to be successful.
5.1.5 The impact on Knowledge Creation

Even though the Lean philosophy commonly focuses more on how to eliminate waste, it is of interest
to think about how the value adding share of the activities at an R&D department would be affected
by an SDPD approach. What a value adding activity is depends on the specific company, but Ward
(2007) analogous to Scania (2010) emphasizes knowledge creation as one of the value streams in the
product development process.
The three types of knowledge learning described by Ward in section 3.2.1, Integration learning,
Innovation learning and Feasibility learning can all be enhanced with help of simulations in different
ways. Simulations performed by design engineers will create a stronger integration between design
groups and simulation groups and therefore contribute to Integration learning. These types of
simulations do also enhance the Innovation learning for the design engineer since the case study
showed that the biggest advantage for a design engineer using GAS in CATIA was they gained
knowledge about different concepts and solutions. Simulations in general are a helpful tool for
Feasibility learning since it is a good way to get decision support about different functionalities.
One crucial part when striving to enhance the level of knowledge in a field is feedback. Simulations
often have the advantage of provide shorter lead time from question to answer in comparison to
physical testing. Decreased lead time for each learning loop will probably shorten the learning time
for each individual. However, the result itself will not gain the knowledge. It is a significant difference
between transferring data and generating knowledge. Knowledge is about understanding and be
able to do accurate analysis from given data. Therefore cross functional teams and experts in
different fields are needed as a support when transferring data into knowledge.
Ward (2007) highlights that “trade-off curves” as a sufficient information carrier when transferring
knowledge between individuals. When using simulations it is rather easy to change one parameter
and get an illustration of how another parameter is affected. The use of simulations should therefore
facilitate the creation of trade-off curves. However, it will likely be an initial inertia when starting to
increase the number of simulations in the product development process. It is a knowledge threshold
that each individual that getting in touch with simulations have to overcome in order to be able to
69
enhance its knowledge. Otherwise the result of the simulations will continue to be confusing data
and SDPD cannot contribute to the knowledge creation.
5.2 Preconditions for Simulation Driven Product Development at Scania

Based on the theory and results from case studies there are some different preconditions to reach a
successful implementation of SDPD at Scania. Even though some of the statements and discussion
are quite company specific it should be possible to generalize findings to a level that is accurate for a
company in the automotive industry that currently use a lot of physical testing within its
development process. Areas such as knowledge and experience, cross functional interaction, and
integration in the product development process are all preconditions that likely are present at more
companies than Scania.
5.2.1 Knowledge and Experience

As mentioned both in the theory framework and results from case studies knowledge and experience
are crucial to be successful with a SDPD implementation. Knowledge within simulation and analysis
will be necessary in order to create accurate models for simulation and justify if the result seems to
be accurate. However, to be able to do accurate simulation models the knowledge within physical
testing is crucial and therefore knowledge and experience are important factors for a SDPD
implementation in most of the different functions in the product development process.
5.2.1.1 Competence within Simulation and Analysis
According to both theory and benchmarking cases, a key success factor with SDPD is to have a deep
core competence about simulation methods and use it to improve and support the simulation
processes. Experts in simulations should not necessarily do the simulations themselves. Scania has
today many engineers already working with simulation and analysis. They have also been involved
with calculations as FEM and CFD many years back and therefore Scania has a fairly long history
when it comes to simulation and analysis. The existing competence is therefore a valuable resource if
a SDPD approach will be implemented. However, the simulation groups at Scania are already choked
and it is a frequent queue of simulation tasks idling for each group. Some groups therefore do not
have sufficient time to develop its simulation models or create a work process for the simulation
tasks which would probably be needed to fulfill an implementation of SDPD.
One way to manage the choked simulation groups and get available time for method and model
development is to try to eliminate the tasks that the simulation engineers are over classified to spend
their time on. Some of the simulations might be executed by the design engineer itself. However, if
more employees, e.g. the design engineers, than only the specialists are going to perform the
simulations, enhanced simulation competence is needed at the design groups. The close
collaboration between a design and a simulation engineer is mentioned by both parties as a
precondition when design engineers try to do simple simulation by their own. To facilitate this
knowledge it could for instance be reasonable to have a simulation engineer geographically located
at each design group or at least a specific engineer responsible for each design group.
One issue with decentralization of simulation engineers is that the individual separated from the
simulation group will lose its depth in competence. One way to solve this, analogous to the Lean
approach, is to arrange some kind of rotation there the simulation engineer is located at a design
group just for limited time period and when it has come to its end it will be replaced by a colleague.
To handle the paradox of increasing core competence and broad the knowledge simultaneously Saab
Automobile used a mixture of a decentralized and centralized simulation groups, by locate the CAE
70
decentralized, but they were tight connected to a centralized group which was responsible to
develop the models and spread knowledge to each individual. In Lean, general work rotation is a tool
used to gain knowledge.
5.2.1.2 Knowledge and Experience about Physical Testing
To be able to simulate more areas in product development, new methods and tools usually needs to
be developed. In order to create new reliable simulation methods, knowledge from physical testing
labs are used. When Saab Automobile implemented SDPD they used Road-to-Lab-to-Math, RLM,
projects and they had to start to transfer signals from road testing to physical testing in labs. Since
Scania is leading within dynamic strength tests and has a long history with physical lab testing, it has
already done half the journey of RLM project by going from road testing to lab testing. Therefore,
Scania has the possibility to take advantage of their knowledge in the physical testing labs to create
mathematical simulation models, the other half of the RLM-projects. When introducing SDPD
physical testing is not going to be eliminated, it will be necessary to advantage of benefits physical
testing has, e.g. discover unexpected cracks, measure loads in specific areas and use it to get a more
accurate simulation model. Physical testing is also needed to evaluate the new simulation methods
and therefore it will still be a need of engineers within the area of physical testing, even though the
tests might change a little bit compared to the current situation; test engineers might be used to
develop new simulation methods and are needed to do validating tests to verify the simulation
models.
5.2.2 Concurrent Engineering

A positive side effect that comes with SDPD is the faster and more frequent interactions between
different departments within the company, e.g. the interference analysis at Boeing, see 3.3.1.4
Interference Analysis. The communication barriers will and has to be reduced in order for SDPD to
work properly. This section will describe and analyze the importance of interaction between different
groups to make SDPD function better.
5.2.2.1 Interaction between Design and Simulation
Since the roles within product development usually consist of a design engineer who does the design
and 3D geometry and different types of simulation engineers who perform the different types of
simulations a stronger interaction is required between them. The first reason is that the transfer of
geometry and information often needs some type of communication since it is based on both the
demands required to perform the simulation, but also the existing data the design engineer already
has. Depending on the PDM/PLM-system, the design engineer also need to inform the simulation
engineer which model should be used and which one that is most up to date. If there would be a
possibility to express maturity levels of the models and if the product development projects had clear
synchronization points, this communication demand would probably vanish.
The other reason is the transfer of the result of a simulation. To just deliver a “pass or fail” answer
will not help the design engineer to improve the design or learn anything about its functionality.
Therefore more simulation jobs usually increase the communication between design groups and
simulation groups since they need to interact more with each other. In order to lower the
communication barriers and to enhance knowledge in each other’s areas, the groups should
preferably be geographically located close to each other as the most groups are at Scania today or as
mention earlier, see 5.2.1.2 Knowledge and Experience about Physical Testing, integrate a simulation
engineer within a design group.
71
The interaction between the simulation groups and design groups will and need to enhance when
some simulations are performed by a design engineer. Since the design engineer has to learn more
about the method of the simulation and how to analyze the results, it will also gain a better
understanding for simulations in general and is able to have a more deepen discussion with the
simulation engineer about more advanced simulations.
5.2.2.2 Interaction between Simulation and Physical Testing
As mention earlier regarding the RLM projects performed at Saab, enhanced or new simulation
methods usually arise from methods and measurement from physical testing. Therefore a strong
connection between simulation groups and physical testing groups is necessary to improve and
extend current simulation methods and tools. Scania has today no clear connection here either
geographically or through the organization. If simulations will increase and the employees at physical
testing will slowly decrease without any clear interaction between the areas, the knowledge about
physical testing might vanish instead of being integrated with the simulation models.
A common process for virtual and physical testing with an obvious and clear connection between the
tests regarding model, calculations and result is needed to enhance the quality of the simulations
and not lose the knowledge Scania has within physical testing.
5.2.3 Simulations in the Product Development Process

Today, there is no clear documentation or instructions about where and how simulations are used in
the product development projects. Though, there is clear documentation about physical testing and
consideration when to include physical testing in the project plans. Since physical test vehicles
require a lot of administrational work and planning, these can sometimes control the time plan for
the rest of the project and is a clear and obvious milestone compared to virtual testing which today
usually does not even have a milestone in the projects at Scania.
When prioritizes are needed due to time pressure, design engineers tend to focus on ordering
prototypes to physical testing earlier instead of complete a simulation job. This is probably done
since the physical test vehicles have clear deadlines and simulation results can sometimes not even
be required. Here can a cadence with simulation loops and synchronization points in the project give
a greater support for simulations in the product development process. It would also affect the design
engineer to prioritize uploading accurate models more frequently. A maturity choice available in GEO
would simplify continuous uploading for the design engineer.
Going through available documentation and materials about the product development process for
design engineers the demands regarding simulation and calculations in an early stage are absent. If a
simulation job is ordered or not for a design is usually more dependent on the interest for
simulations within each and every design engineer and if the design group leader has a background
within simulation and analysis or not rather than the actual need for a simulation on that specific
design.
5.2.3.1 Simulations performed by Design Engineers
Today at Scania, design engineer can do some FEM simulation by themselves in CATIA using GAS. This
is a helpful tool to increase the amount of simulations without necessarily increase the number of
simulation engineers. The simulation engineer can instead of doing the simulations themselves act
more of a support helping the design engineers perform accurate simulations. The GAS application is
not used by all the design engineers, but only by them who have found it interesting and therefore it
is mainly used by younger employees with little experience as a design engineer. The reason why this
72
group is overrepresented among GAS users is probably because they have a stronger demand for
learning about the robustness in different designs as fast as possible and will not have time to wait
for feedback from an ordered simulation job or, even longer, a physical test.
5.2.3.2 Front-Loaded Projects
It is crucial when in the product development process simulations are performed. It should be an
activity that is scheduled and planed very early in the projects, before any physical testing. According
to lean product development and set based concurrent engineering, it is desirable to use simulation
results to evaluate concepts and letting the analysis from the simulations be essential when taking
important decisions about which concept should proceed and not. Since virtual testing has a very
short lead time compared to physical, it is very smooth to use early in the projects. It will not push
the activities closer to SOP as physical testing would. It also makes it possible to do more iterations
per concept or idea, i.e. more work, compared to physical testing, which means that using a SDPD by
performing many simulations early in the projects will front-load the projects.
5.2.4 Incentives for an Implementation

Incentives can be a driving factor to make a bigger change, e.g. a SDPD implementation, within a
larger organization. This section aims to investigate the plausible incentives for an implementation of
SDPD both at Scania and within the automotive industry in general.
5.2.4.1 Incentives to reduce Lead Time for Product Development
The research and development department at Scania has newly started the challenge ”Double
Output – Half time” which should reduce waste at R&D in order to either reduce lead time, double
the outcome or both. The goal is not clear and can be interpreted in many ways, but the main
purpose is to develop new products faster. This challenge is therefore a good reason and incentive to
use a SDPD. Instead of letting the time demanding physical tests and generation prototype trucks be
the base of the time plan, simulation loops can be the backbone of the process. Since a simulation
loop takes a lot less time and resources it will shorten the overall project time and physical testing
should only be used in the end of the projects to verify the simulation models. The lead time
reduction is also the main cause of the SDPD approach at AB Volvo. Since the two companies are
present within the same business segment demanded lead time reduction is a feasible incentive to
start a SDPD journey at Scania.
5.2.4.2 Economic Aspects of a Simulation Driven Product Development Implementation
As described in the theoretical framework, see 3.3.2 Economic Aspects of implementing Simulation
Driven , as well as the results from case studies, see 4.2.1 Cost Rationalization by using Simulations at
Saab Automobile AB, cost reduction is a common documented outcome of a SDPD implementation.
For Saab, it was the driving force for the SDPD implementation. For a company that currently uses a
lot of physical testing in its product development process, it is possible savings in an implementation
of SDPD and it can therefore be a good incentive for usage of SDPD.
An estimation in of how large the amount of savings a SDPD implementation should generate is
rather complex due to that a lot of individuals, departments and resources are connected to physical
testing in one way or another. Some interviewees taking part in the case study at Scania emphasize
for instance that a purchaser spend 50-70 % of its time on sourcing prototypes. Even though the time
share is not confirmed a lot of time at the purchasing department will be related to prototypes.
Testing engineer is another role tightly related to physical testing and the costs to sustain the test
facilities is also due to physical testing. Therefore it is very difficult to come up with an overall
73
potential saving. At the same time an SDPD implementation does not necessarily mean reduced
employees within physical testing. As mention in section 5.2.1.2 knowledge and experience about
physical testing should be used for the implementation and to develop the simulation methods.
However, as a simplification and quite conservative calculation the cost for prototypes can be used.
Volvo emphasizes that the design engineers are supposed to choose the most time and cost efficient
alternative when evaluating the design of a new product. In the normal state of the development
phase, virtual testing is more rational to use in comparison to physical testing. However, it is still
areas then physical testing outperform virtual testing, see 4.2.3 Simulation Development Loops at AB
Volvo, and in this cases physical testing should still be used even though it is a target of zero
prototyping. To be able to make such decisions the design engineer has to have a high cost
consciousness of what kind of expenses each test method correspond to. This is also connected to
the Lean theory regarding cost consciousness in section 3.2.4 If it is a lack of this cost consciousness,
it is likely that the cost rationalization is not being realized. Instead the category of waste referred to
as hand-off, see section 3.2.2.2 will increase in the product development process. Therefore, it is of
great importance to put attention to and sustain cost consciousness among design engineers to be
able to use economical factors as an incentive for a SDPD implementation.
5.2.4.3 Environmental Consciousness
The reduction of prototypes in a SDPD in comparison to a traditional approach will not only affect the
cost. It also has an impact on the environmental image and footprint of the company. Therefore
environmental consciousness and sustainable development should be possible incentives for a
change towards SDPD.
Jaguar Land Rover has used the environmental aspect as one of the main reasons to have the target
Zero Prototyping. As the automobile industry has a bigger pressure today to be profiled as
environmental friendly, not only the CO²-emissions, but also the environmental foot print from the
product development and production can also affect the association between the automobile brand
and environmental awareness.
However, this incentive is unsupported by the theory as well as the other company cases. Though,
the environmental consciousness has increased a lot during the last years and the articles the
theoretical framework is based on are a couple of years old. Several companies within the
automotive industry has the environment as a core value, e.g. AB Volvo and Scania, and that is
another reason why environmental consciousness should be a possible incentive for a change
towards SDPD.
5.2.5 Trust to Simulation Results

A common factor for both AB Volvo and Saab automobile is the reliability they have to simulations
and how they interpretive the results regarding accuracy. When the simulation results are not
consistent with the results from physical testing, the measuring methods are questioned before they
doubt the simulation models. A former employee at Saab automobile mentions that simulations are
used in the very beginning of a project to convince other about a concept or idea. The trust to
simulations is widely spread throughout the organization and therefore the simulation results have a
big importance when taking decisions.
Since Scania has a long history and experience of physical testing, its employees tend to rely and
draw conclusions on the results from physical testing more often than simulations and then the
results differ, the simulation methods are usually questioned over physical testing. If the simulation
74
results are not given the attention it should and is never used ahead of results from physical testing,
there is now reason to perform the simulations. At the same way the suspicion for simulations and
virtual testing will prevent SDPD. The over confidence for physical testing is dangerous for the quality
of the products. One broken component in a physical dynamic strength test will not guarantee that
all components will break and vice versa due to the distribution in a physical population of
components. Therefore, a single physical test should not have a greater importance than a
simulation result with several iterations behind it. They should instead support each other trying to
explain the whole scenario, e.g. predict where cracks can occur and explain why and how likely it will
be.
The trust is probably a success factor for the implementation and therefore a broaden trust to
simulation results needs to be established if SDPD are going to be implemented. Without trust each
project needs to overtake inertia of doubt and unnecessary physical testing will still be performed in
order to secure the result. According to the theory chapter trust can be gained by success stories.
Therefore the pilot project for the SDPD implementation is crucial for creating trust to simulations.
The initial project should be simpler than the tools have potential for in order to minimize the risk of
failure. A failure at an early stage will probably decrease the trust to simulations and therefore
increase the threshold for starting a journey towards SDPD.
75
76
6 CONCLUSIONS
This chapter comprises the output to the research questions introduced in chapter 1 as well as Scania specific
recommendations. The chapter is divided into two sections; the first concluding the result of RQ1, RQ2, RQ3
while the second section conclude the recommendations.
6.1 Result of the Research Questions

The applicability of implementing SDPD in a Lean Enterprise is investigated in the analysis of the
thesis and is according to it rather high. The effects of a SDPD implementation in a Lean environment
will primarily affect three different main categories which is the output of RQ1 and are presented in
the box below.
APPLICABILITY OF COMBINING SIMULATION DRIVEN PRODUCT

DEVELOPMENT AND LEAN PRODUCT DEVELOPMENT
• It is a consistency between SDPD and the Lean Principles: SDPD support or have a
positive impact on several of the 13 principles describing Lean Product Development.
In several cases it can be a tool to sustain the principles. However, SDPD in itself will
not create a Lean product development process; it is closely connected to how the
implementation is done and in what way SDPD is used.
• SDPD reduces waste: SDPD implemented in perspicacious manner will reduce the
amount of waste in the development process in several ways. The waste reduction
potential is specially clear in the categories Scatter and Hand-Off, but it will most
likely reduce all different types of waste more or less. Although, it is a risk that SDPD
will cause waste; primarily related to “doing more than necessary” if it is
implemented in the wrong way.
• SDPD is a precondition for using Set-Based Concurrent Engineering: The literature

related to Lean Product Development consistent highlights the importance of basing
decisions on facts and working with several concepts in parallel. SDPD is a
precondition to be able to follow this method in resource efficient and feasible way.
Without SDPD, set-based concurrent engineer would be very costly, time consuming
and therefore an insufficient methodology to use in a Lean environment
77
CHAPTER 6 – CONCLUSIONS
The preconditions for SDPD can be found in the analysis chapter and are divided in six areas
described further in the box below. Even though the analysis had a focus on preconditions for Scania,
these result be can applicable in a more general perspective within the mechanical engineering
business.
PRECONDITIONS FOR SIMULATION DRIVEN PRODUCT DEVELOPMENT

• Incentives from Management: Incentives are needed to create a more clear
understanding of the importance of the simulations to achieve an increased usage of
simulations among the research and development organization. Therefore, simulation
driven product development should be supported by economical, environmental and
lead time reducing incentives.
• Knowledge and Experience within physical testing and simulation: Simulation Driven
Product Development requires a high level of knowledge and experience within both
virtual and physical testing in order to develop new advanced methods that can be
used to evaluate complex technical functions on the product. The new methods for
virtual testing should grow out of experience and knowledge of physical testing.
Physical testing is also used to validate virtual testing and simulation methods.
• Interaction between design, physical testing and simulation: Simulation driven

product development require frequent interaction between design and simulation
groups in order to always have accurate virtual models for the simulation and
continuously aware the design engineer about simulation results. It also require a
clear interaction in the process between simulation groups and physical testing so
they can have the use of each other’s results and experience as described in the
paragraph above.
• Clear and front-loaded development process: If a product development process

should be simulation driven, it needs to be clear when and how virtual models should
be created, distributed and simulated. Analogous to Lean principles, the process
should be front-loaded and simulations will have bigger effect on quality, lead time
and cost if they are performed earlier in the product development process.
• Synchronization and maturity levels: Simulation driven product development will

require full-vehicle virtual prototypes and these can be more easily achieved with a
PDM/PLM-system containing maturity levels for the geometry models.
Synchronization points in the project plan for full-vehicle models will also ease the
possibility to perform full-vehicle simulations efficiently.
• Courage and trust to simulations: Simulation driven product development requires a

brave organization willing to trust and daring to believe simulation results. This
courage is required both in management, but also widely spread among engineers in
different areas.
78
The efficiency of a simulation driven product development process can be increased when it is
combined with simulations performed by design engineer. The box below presents the five most
important impacts design engineer simulations have on a product development process.
IMPACTS OF USING DESIGN ENGINEER SIMULATIONS

• Enhanced understanding for physical properties of the component: By using
simulations, the design engineer achieves the opportunity to understand the strength
and limitations of the design more easily than when it receives an answer from
someone else. The enhanced understanding creates an opportunity to achieve more
optimum solutions.
• Enable a larger number of simulation iterations: Since the time required for each
simulation loop is reduced from approximately two weeks to a couple of hours it is
possible to do more simulations and therefore reach a more optimum design than if
simulations are performed by a simulation group.
• Increased integration between design and simulation groups: To be successful with

the simulations the design engineer needs support from simulation engineers. This
collaboration will increase the understanding for each other’s circumstances. It will
probably also reduce the communication barriers and enhance the knowledge
transfer between the groups.
• Increased possibility of front-loading projects: Simulations performed by design

engineers can be involved at an earlier stage in the development phase than regular
simulations. The design engineer can for instance do simulations during the concept
phase to investigate some general ideas with a very low maturity of the CAD
geometry.
• Decreased product development lead-time and increased resource efficiency: For each
simulation a design engineer performs by itself instead of hand-over to another
person idle time and waiting are reduced as well as start up times for the individual
who otherwise are going to perform the simulation.
79
6.2 Recommendations to Scania

With support from the analysis and the conclusions some specific recommendations to Scania are
formulated regarding simulation driven product development. These recommendations are not as
generable as the result since they depend on the current state at the company.
SCANIA SPECIFIC RECOMMENDATIONS REGARDING SIMUALTION DRIVEN

PRODUCT DEVELOPMENT
• Establish targets for reducing prototyping: Today, there is no clear incentive regarding
reducing prototypes. A vision or goal would help employees to understand the
importance of the amount of prototypes used and the engineers can more easily
prioritize simulations. A reasonable target in a short time horizon might be the 35 %
reduction mentioned earlier in the thesis. In order to be consistent regarding
incentives it is important to connect other established incentives, such as Double
Output – Half time, with simulation driven product development.
• More clear product development process regarding simulation and analysis: Today,
there is no documentation when or how simulation and analysis should be used in the
product development process. This should be established, especially from the
perspective of the design engineer. Here the feedback process between physical
testing and simulation, mentioned in the paragraph below, should also be included.
• Secure a more clear feedback process between physical testing and simulation:
Without a clear demand for feedback between the physical testing groups and the
simulation groups in the product development the connection between the groups
will not vanish. How the feedback should be performed needs to be defined since it is
crucial for the interaction between the two areas and the validation of the simulation
methods.
• Invest in research regarding new simulation methods for dynamic strength: The
biggest gap between physical testing and simulations for Scania is in the area dynamic
strength and durability. Since it is one of the most important factors for Scania and its
products the main research focus for simulations in the nearest future should be
within this area.
• Systematic simulation method development to fulfill the testing demands: Scania will
continuously need simulation method development projects where the testing
demands are moved from Road to Lab to Math (RLM-projects). This should be made
to be able to reduce physical prototypes.
• A general support structure within the organization regarding simulations performed

by design engineers: Simulations for design engineers should be supported with help
of simulation engineers. There should be a support function outspreaded in the
organization closely located to the design engineers.
• Require geometries and models more frequently: A demand for getting accurate
models earlier in the projects is a presumption to be able to perform sub-system or
full-vehicle simulations. This requires a possibility to express maturity and to have a
tact with synchronization gates in the product development projects.
80
7 FURTHER STUDIES
During the study some areas there further research would be appreciated has been detected. The areas presented
in this chapter is beyond the limitations for the study and would be of interest to investigate further in order to
be able to do a successful implementation of simulation driven product development.
Measure the amount of rework: Unnecessary rework is referred to as a substantial kind of waste.
Since rework is required when other kind of wastes, e.g. wishful thinking, is present it might be a
good KPI to measure quality during the product development process. Therefore, it would be of
interest to consider how rework can be clearly measured and defined.
Examine how to arrange a Road-to-Lab-to-Math journey: In order to achieve a high accuracy in
simulations they have to reflect the reality. In order to success with it, RLM-projects will be
necessary. Therefore the RLM arrangement has to be investigated in order to be able to implement
simulation driven product development.
Increase the cross functionality of the scope: Simulation tools are used in several other departments
as well. This study have a focus on research and development, but it would be of interest to get a
more holistic view and have the entire organization including marketing, production and purchasing
in order to develop a common platform for simulation driven product development.
Investigate the possibility to handle maturity levels of items and cadence in the IT-systems: Best
practice companies within simulation driven product development have implemented IT-systems
that can handle levels of CAD geometry maturity and has a cadence in its process. It would be of
interest to study how this arrangement could look like at Scania CV AB.
Perform a more detailed economic calculation: The economic impact of a simulation driven product
development implementation has to be investigated more in detail. This study only comprises an
example of reduction of prototypes. Other possible savings should also be investigated. It is also of
interest to investigate the required cost for investment in both hardware and software as well as
competence development in order to manage a change towards simulation driven product
development.
Investigate the effects on lead time: The positive effect on lead times by implementing simulation
driven product development is well documented, but it would be of interest to study how much the
lead time can be decreased and what effects it would generate.
Available information from subcontractor: A topic that will affect the complicity of an
implementation is what kind of information the subcontractors offer. The available CAD geometries
and simulations should be investigated. It is also interesting to study the requirements on
subcontractors. It is maybe possible to force the subcontractors to offer results from simulations as a
complementary to protocols from physical testing.
Investigate sufficient share between simulations performed by design engineer and experts: In order
to keep the accuracy good enough on the simulations the allocation of simulations should be studied.
Some kinds of simulations are suitable for the design engineer to perform while others have to be
performed by experts. It is therefore a need of developing some kind of guideline for in what way
simulations should be executed.
81
CHAPTER 7 – FURTHER STUDIES
Investigate the maturity and accuracy of different simulation models: As a subpart or a first step of a
RLM journey the accuracy of current simulation models should be investigated in order to find areas
there it is a need of enhanced model development to be successful with a future RLM journey.
Development of detailed implementation guidelines: The result of this study only comprises some
general recommendations of actions in order to take some steps towards a simulation driven product
development. Therefore it is a need of develop a more detailed framework for an implementation of
simulation driven product development
82
REFERENCES
Aberdeen Group, 2006. Simulation Driven Design Benchmark Report Getting It Right the First Time.
[Online] Aberdeen Group Available at: http://www.aberdeen.com/Aberdeen-
Library/3591/BM_Simulation_driven_Design_3591.aspx [Accessed 14 March 2012].
Adams, V., 2006. How To Manage Finite Elment Analysis in the Design Process. Glasgow: NAFEMS Ltd.
Andersen, H., 1994. Vetenskapsteori och metodllära. Lund: Studentlitteratur.
Anderson, J.D., 2009. Govering Equations of Fluid Dynmaics. Thrid edition ed. Berlin: Spinger-Verlag.
Ansys Inc., 2011. Measuring Return on Investment. [Online] Ansys Inc Available at:
http://www.ansys.com/staticassets/ANSYS/staticassets/businessinitiative/exec/Measuring-Return-
on-Investment.pdf [Accessed 14 March 2012].
Bergman, B. & Klefsjö, B., 2010. Quality from Customer Needs to Customer Satisfaction. Third edition
ed. Studentlitteratur AB.
Björklund, M. & Paulsson, U., 2007. Seminarieboken - att skriva presentera och opponera. Lund:
Studentlitteratur.
Boon, D., 2009. Virtual testing of chassis structures. [Online] Available at:
http://webcache.googleusercontent.com/search?q=cache:QMYE0RuZdagJ:www.twi.co.uk/EasysiteW
eb/getresource.axd%3FAssetID%3D17673%26type%3DFull%26servicetype%3DAttachment+JLR+%22
virtual+testing%22&cd=1&hl=sv&ct=clnk [Accessed 14 Mars 2012].
Brantefors, O. & Gadman, F., 2007. Gränssnitt: Definiering, hantering & delning - En utredning av
Scanias möjligheter till ett aktivt utnyttjande av gränssnitt vid produktutvekling. Masters Thesis.
Stockholm: KTH Royal Institute of Technology Industrill teknik och management.
Bryman, A. & Bell, E., 2011. Business Research Methods. Third edition ed. New York: Oxford
university Press.
CIMA, 2009. Jaguar Land Rover Case studies. [Online] Chartered Institute of Management
Accountants Available at:
http://www.cimaglobal.com/Documents/Thought_leadership_docs/Sustainability%20and%20Climat
e%20Change/jaguarcasestudydec09.pdf [Accessed 14 Mars 2012].
Creswell, J.W., 2003. Research design: Qualitative, Quantitative and Mixed Methods Approaches.
Second edition ed. Thousand Oaks: Sage.
Dassault Systèmes, 2012. Virtual ergonomics - ergonomic evaluation. [Online] Available at:
http://www.3ds.com/products/delmia/solutions/human-modeling/portfolio/ergonomics-evaluation/
[Accessed 29 March 2012].
Davidsson, B. & Patel, R., 2003. Forskningsmetodikens grunder. Lund: Studentlitteratur.
83
Diesel & Gas Turbine Publications , 2003. Audi plans digital vehicle assembly with Delmia solutions.
[Online] Available at:
http://findarticles.com/p/articles/mi_m3012/is_5_183/ai_101939345/?tag=content;col1 [Accessed
29 March 2012].
Gripenberg, P., 2006. Kriterier för bedömning av forskning: Validitet, reliabilitiet och replikerbarhet
inom kvantitativ och kvalitativ metodlära. [Online] Instiitutionen för företagsledning och organisation
Available at: http://brunnen.shh.fi/portals/studymaterial/2008-
2009/helsingfors/foretagsledningochorganisation/2235/material/useful/Validitetreliabilitet2.pdf
[Accessed 17 April 2012].
Hirsch, C., 2007. Numerical computation of internal and external flows: The fundamentals of
computational fluid dynamics. second edition ed. Oxford: Elsevier.
Holmdahl, L., 2010. Lean Product Development på svenska. Göteborg.
Holme, I.M. & Krohn, S.B., 1997. Forskningsmetodik - Om kvalitativa och kvantitativa metoder.
Second Edition ed. Lund: Studentlitteratur AB.
Honeywill, T., 2008. Analyse this. Dale earnhardt engineering, September. pp.37-40.
Hutton, D., 2005. Fundamentals of Finite Element Analysis. New York: McGraw-Hill.
IDC, 2011. IDC Announces First-REcipients of HPC Innovation Excellence Award. [Online] Available at:
http://www.idc.com/getdoc.jsp?containerId=prUS22897711 [Accessed 15 March 2012].
Jackson, C., 2007. NAFEMS Webinar series. [Online] NAFEMS Ltd. (1) Available at:
http://www.nafems.org/events/nafems/2007/SDDFindings/ [Accessed 14 13 2012].
Jacobs, L. & Herbig, P., 1998. Japanese product development strategies. Journal of Business &
Industrial Marketing, 13(2), pp.132-54.
Jaguar, 2012. Waste reduction Through Virtual Testing. [Online] Available at:
https://jaguar.credit360.com/jaguar/site/page3.acds?context=980370&instanceid=980371
[Accessed 3 April 2012].
Johannesson, H., Persson, J.-G. & Pettersson, D., 2004. Produktutveckling - effektva metoder för
konstuktion och design. First edition ed. Stockholm: Liber AB.
Johannesson, H., Persson, J.-G. & Pettersson, D., 2005. Produktutveckling - effektva metoder för
konstuktion och design. 1st ed. Stockholm: Liber AB.
Karlsson, C., 2006. Analys av brott vid krocksimulering av lastbil. Master Thesis. Luleå: Luleå tekniska
universitet Institutionen för tillämpad fysik, maskin- och materialteknik.
Karlsson, E., 2006. En åkandesimuleringsmodell i LS-DYNA. Master Thesis. Lund: Lunds tekniska
högskola Byggnadsmekanik.
Kennedy, M.N., 2003. Product Development for the Lean Enterprise - Why Toyota's System is Four
Times More Productive and How You Can Implement It. First Printing ed. Virginia: The Oaklea Press.
84
Liker, J.K., 2004. The Toyota Way. New York: McGraw Hill.
Liker, J., 2011. Toyota under Fire: How Toyota Faced the Challenges of the Recall and the Recession to
Come out Stronger. McGraw-Hill.
Liker, J.K. & Morgan, J.M., 2006. The Toyota Way in Services: The Case of Lean Product Development.
Academy of Management, 20(2), pp.5-21.
Loman Strinnholm, R., 2012. Development of Methodology for Simulation Driven Design. Mater
Thesis. Göterborg: Chalmers University of Technology, Department of Product and Production
development.
Morganed, J. & Liker, J.K., 2006. The Toyota Product Development System. Portland: Productivity
Press.
Morgan, J.M. & Liker, J.K., 2006. The Toyota Product Development System. First Edition ed. New York:
Productivity Press.
Normile, D., 2001. Toyota takes auto assembly digital. Design News, 5 February.
Oosterwal, D.P., 2010. The Lean Machine. New York: AMACOM.
Raphael, B. & Smith, I.F.C., 2003. Fundamentals of Computer-Aided Engineering. West Sussex: John
Wiley.
Reinerstein, D.G., 2009. The Principles of Product Development Flow. First Edition ed. Redondo
Beach, California: Celeritas Publishing.
Scania CV AB, 2010. R&D Factory. Internal Company Report. Sweden: Scania CV AB R&D Factory
Office.
Scania CV AB, 2011. PRODUCT DEVELOPMENT PROCESS. Internal Company Report. Sweden: Scania
CV AB Project Office YP.
Scania CV AB, 2011. Scania in brief. [Online] Available at: http://scania.com/scania-group/scania-in-

brief/ [Accessed 10 Januari 2012].
Scania Inline, 2011. Product development process: Scania Inline. [Online] Project Office YP (2)
Available at: http://inline.scania.com/oliver_upload/upl393619-STD4303en.pdf [Accessed 10 Januari
2012].
Shigley, J.E., 1963. Mechanical Engineering Design. First edition ed. New York: McGraw-Hill.
Sjödin, T., 2010. Saab 9-5 Virtual Development Without Hardware Prototypes. In NAYFEMS Nordic
Conference., 2010.
Sobek, D.K., Ward, A. & Liker, J.K., 1999. Toyta's Principles of Set-Based Concurrent Engineering.
Sloan Management Interview, 40(2), pp.67-83.
Thomke, S. & Fujimoto, T., 2000. The effect of "Front-Loading" Problem-Solving on Product
Development Performance. Elsevier, 17, pp.128-42.
85
Toyota, 2003. Taking the initiative. Annual report. Toyota motor company.
Wallen, G., 1996. Vetenskapsteori och forskningsmetodik. Second edition ed. Lund: Studentlitteratur.
Ward, A., 2002. The Lean development Skills Book. michigan: Dollar Bill Books.
Ward, A., 2007. Lean Product and Process Development. 1st ed. Cambridge: The Lean Enterprise
Institute, Inc.
Ward, A., Liker, J.K., Cistiano, J.J. & Sobek II, D.K., 1995. The Second Toyota Paradox: How Delaying
Decisions Can Make Better Cars Faster. MIT Sloan management Review, 36(3), pp.43-61.
Wernsten, T. & Hanna, S., 2012. Informationshantering med kravsättning i 3D-baserade arbetsflöden.
Masters Thesis. Stockholm: KTH Royal Institute of Technology Industriell teknik och management.
Westling, J., 2001. Matematiska modeller ersätter prototyper. NyTeknik, 15 August.
Womack, J. & Jones, D., 1991. The machine that changed the world: The story of lean production.
New York: Harper Perennial.
Yin, R.K., 2008. Case Study Research Design and Methods. Fourth edition ed. Thousand Oaks.
86
APPENDIX A - NOMENCLATURE
ABACUS A suite of software from the software provider Dassualt Systèmes. For
instance used to perform finite element analysis.
CAD Computer-Aided Design
CAE Computer-Aided Engineering
CATIA Computer Aided Threedimensional Interactive Application - A CAD
software from the software provider Dassault Systèmes
CFD Computer Fuild Dynamics
Chalmers Chalmers University of Technology
COD Cost of Delay
DELMIA Digital Enterprise Lean Manufacturing Interactive Application – A
software from Dassualt Systèms used for digital manufacturing
ECO Engineering Change Order
EURO-NCAP The European New Car Assessment Program - a European car safety
performance assessment program
FEA Finite Element Analysis
FEM Finite Element Method
GAS Generative Assembly Structural Analysis – an integrated tool that allows
designer engineers to analyze assemblies as well as individual parts in
CATIA V5
GEO Abbreviation for geometry assurance and a system used at Scania to
share CAD geometries
GP Geometric Position
GM General Motors
JLR Jaguar Land Rover
KTH KTH Royal Institute of Technology
LOM Level of Maturity
PA Scania abbreviation for Testing Request
PD Product Development
PDM Product Data Management
PLM Product Lifecycle Management
R&D Research and Development
I
R&D Factory Adaptation of Scania Production System for the Research and
Development environment
RLM Road-to-Lab-to-Math – Going from physical to virtual testing
ROI Return of investment
RQ Research Question
Saab Saab Automobile AB
SBCE Set-Based Concurrent Engineering
Scania Scania CV AB
SDPD Simulation Driven Product Development
SDD Simulation Driven Design
SPS Scania Production System
Volvo AB Volvo
II
APPENDIX B – SBCE FRAMEWORK
This appendix comprises a framework for Set-Based Concurrent Engineering developed by Sobek,
Ward and Liker 1999 and has its origin in the article Toyota's Principles of Set-Based Concurrent
Engineering.
SBCE FRAMEWORK ADAPTED FROM SOBEK ET AL. (1999)

Principle 1: Map the design: Aiming to support the team to come up with different
possibilities to fulfill a need of a function. Adapt
• Define feasible regions: This is a cross functional activity where each single
department involved independently and in parallel tries to come up with constraints
to the new design in its own environment. Each department check in designers’
checklist in order to keep the knowledge from previous projects. The checklists at
Toyota is not a long list, instead it contains holistic guidelines of how to perform more
in general than in detail.
• Explore trade-offs by designing multiple alternatives: The key of this implementation

principle is to investigate trade-offs with the possible multiply subsystems for
solutions. Since the decisions should be based on fact prototypes, simulations and
analysis should be used for investigation. Best guess is not good enough.
• Communicate sets of possibilities: In this session the sets of possible concepts is

communicated. Standard design matrices are used both to communicate the
solutions and to give feedback.
Principle 2: Integrate by intersection: In this phase, design teams starts to integrate

different subsystems in order to identify solutions that is workable for everyone in the
cross functional team.
• Look for Intersections of Feasible sets: The team is looking for intersections in
interfaces between different possible subsystems to be able to combine them. It is
important to have the overall system performance in mind and not just the subsystem
that is in focus for the moment.
• Impose Minimum Constraint: Toyota’s approach is minimum constraints needed at

the time. Flexibility in a late stage in the development process makes it possible to be
agile to the customers need. Often, the need is unclear when the project is launched,
but will be more and clearer when the market launch comes closer.
• Seek Conceptual robustness: Strive to create concepts that are as robust to their
environment as possible. It includes wear, weather and other external impacts as the
surrounding layout and interaction with other concepts.
III
Principle 3: Establish Feasibility before Commitment: It is of great importance to
understand the overall system before the team is getting too committed to the final
solution.
• Narrow sets gradually when increasing detail: The solutions should gradually be
eliminated due to the limitations in each solution. The narrowing has to happen
sometimes and should be based on knowledge and concept failure, i.e. the concept
that seems to be the best one should not be chosen, the least appropriate one should
be eliminated until it is only one left. The responsibility to eliminate a solution and
narrow should be a decision of the project manager.
• Stay with Sets Once Committed: It is important that the team stays within the
narrowing funnel so all participants can be sure that they do not have to take other
possible changes into consideration.
• Control by Managing Uncertainty at Process Gates: Toyota uses a set of both written
and implied rules to handle uncertainty. One example is that if only one concept is left
when it comes to the point of calculating cost, the narrowing process has been to
rapid and more possible solutions or concepts has to be evaluated.
IV
APPENDIX C – DESIGN ENGINEER QUESTIONS
This appendix comprises the questions used during the qualitative study with design engineers. Since
the interviews were semi-structured, follow up questions might take place during an interview.
- How does the daily work as a design engineer look like? What are your actual work tasks?
- Do you follow any process when you perform your work? How does it affect your work?
- How do you do when designing new components? Do you use any simulation tool such as
GAS by your own? Do you develop several concepts in parallel?
- How do you use simulations and analysis in your work? How do you order simulation tasks?
When in the process do you do it? Do you perform simulations on one or several concepts?
- What are the main differences between virtual and physical testing?
- What share of you time at work do you perform pure design tasks? What does the rest time
comprise?
- What kind of activities in your daily work would you enhance your knowledge as a design
engineer? What share of your work time do you spend on these activities?
- How do you secure “right-from-me” when doing a hand-over to someone else or finish a
project?
- Do you have any possibility to affect the expenses for the methods you use when developing a
new component?
- How much would you estimate the cost to be if your delivery to a project is 30 days late?
V
APPENDIX D – LIST OF DESIGN ENGINEERS
This appendix comprises information about the participants in the case study regarding design
engineers at Scania and has answered been respondents to the questions in Appendix A.
Role at Scania Area Years at Scania
1 Design Engineer Basic Chassis Development 2

2 Design Engineer Bus Frames and Installation 8
3 Design Engineer Cab Body and Suspension 4
4 Design Engineer Axle Design 1
5 Senior Design Engineer Bus Powertrain Components 10
6 Design Engineer Engine Valve System 4
7 Design Engineer Engine Valve System 4,5
Cab Instrument Panel and Driver
8 Design Engineer 1
Unit Control
9 Object Manager Basic Chassis Development 8
10 Consultant (Design Engineer) Bus Powertrain Components 1
11 Design Engineer Engine Intake Components 5
12 Design Engineer Chassis Components 1
14 Senior Design Engineer Engine Composition 6
Cab Instrument Panel and Driver
18 Design Engineer 4
control unit
19 Design Engineer Air and Fuel System Layout 2
20 Design Engineer Manual Gearbox 4
21 Consultant (Design Engineer) Chassis Components 15
22 Design Engineer Bus Steering and Suspension 4
23 Consultant (Design Engineer) Bus Brakes and Auxiliary system 2
24 Consultant (Design Engineer) Chassis Components 1
VI
APPENDIX E – SIMULATION ENGINEER QUESTIONS
This appendix comprises the questions used during the qualitative study with simulation engineers.
Since the interviews were semi-structured follow up questions might take place during an interview.
- Do you follow any method or work process when performing a simulation for a design
engineer? How do those function?
- When does simulation and analysis get involved in a development project? E.g. when a new
bracket is going to be developed.
- How long time does a computation take?
- How do you do to assure that the simulations are good enough i.e. not deliver more accuracy
than is actually needed?
- Do you have a queue of incoming simulations? What is your available simulation capacity?
What is the bottleneck (employees, HW resources or licenses)?
- How does the work load looks like over time? Is it leveled?
- What are the most common design mistake from your perspective?
- Which of these might the design engineer find out by themselves in software such as GAS?
- How large share of your time is spent on these kind of issues?
- Do the mistakes a design engineer does differ dependent on their experience?
- Does the design engineer demand simulations on several concepts in parallel?
- Do you get feedback from physical testing? How is this feedback arranged?
- How large amount of your time is spend on “firefighting” simulations?
- Do you do simulations before, after or in parallel with physical testing?
VII
APPENDIX F – LIST OF SIMULATION ENGINEERS
This appendix comprises information about the participants in the case study regarding simulations
engineering at Scania and is respondents to the questions in Appendix C.
Role Area
1 Head of Dynamics and Acoustics Engine Dynamics and Acoustics

2 Technical Manager Axle Technical Simulation
3 Head of Strength Analysis Engine Strength Analysis
4 Calculation Engineer Engine Strength Analysis
5 Development Engineer Bus Dynamics and Strength Analysis
6 Head of Strength and Crash Analysis Cab Strength and Crash Analysis
7 Head of Fluid Mechanics Fluid Mechanics
8 Technical Manager Fluid Mechanics
9 Head of Dynamics and Strength Analysis Truck Chassis Dynamics and Strength Analysis
VIII
APPENDIX G – LIST OF INTERVIEW PARTICIPANTS
This appendix comprises information about the participants of the unstructured interviewed used to
gather an enhanced understanding as well as information in specific areas when developing the case
framework.
Role Area
1 Head of Load Analysis Truck Chassis Load Analysis

2 Head of Dynamics and Strength Analysis Bus Dynamics and Strength Analysis
3 Head of Frames and Installations Bus Frames and Installations
4 Head of Steering and Suspension Bus Steering and Suspension
5 Method responsible Geometry Assurance
6 Project Coordinator Truck Chassis Business Development
7 Senior Technical Manager Strength Testing
8 Head of Acoustics Truck Chassis Acoustics
9 Head of Crossfunctional Product Information Crossfunctional Product Information
10 Method Developer Project Office
11 Technical Manager Geometry Assurance
12 Head of Geometry Assurance and Testing Geometry Assurance and Testing
13 Consultant Geometry Assurance
14 Controller Truck Chassis Business Development
15 Object Manager After Treatment system
16 Development Engineer Turbocharger
17 Senior Engineer Truck Chassis Durability Testing
18 Senior Engineer Truck Chassis Durability Testing
19 Project Manager Superior Quality
20 Business developer R&D Factory Office
21 Project Coordinator Project Office
Head of Business Development and IT
22 Business Development and IT Infrastructure
Infrastructure
23 Method Developer Simulation Driven Design
24 Head of Product Modelling Methods Product Modelling Methods
IX
Role/Area Company
1 Deparment Manager for CAE Saab Automobile AB

2 Head of Geometry Assurance and DMU AB Volvo
3 Client Executive Dassault Systémes
4 Technical Manager Dassault Systémes
5 Director of Business Development EDR & Medeso AB

Simulation Driven Product Development.

Uploaded by

Copyright:

Available Formats

Simulation Driven Product Development.

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Simulation Driven Product Development.

Uploaded by

Copyright:

Available Formats

Simulation Driven Product Development

How it can be combined with Lean Philosophy to achieve increased

Master of Science Thesis in the Master Degree Program Supply Chain

Department of Technology Management and Economics

Master of Science Thesis

© SARA JOHANSSON & DAVID SÄTTERMAN, 2012

Report No. E 2011:021

Gothenburg, Sweden, 2012

Södertälje, May 2012

Sara Johansson David Sätterman

1.3 Problem Analysis

1.5 Outline of the Thesis

5. RESULTS FROM CASESTUDIES

5. ANALYSIS AND DISCUSSION

2.1 Method Model of the Study

Figure 1 Method model for the Study

2.1.1 Method to find output to RQ1

2.1.2 Method to find output to RQ2 and RQ3

2.2 Data Collection

2.2.1 Document Study

2.3 Reliability and Validity

3.1 Product Development in General

3.1.1 Traditional Product Development Processes and Methodologies

Figure 2 Example of Point-based engineering approach

Figure 3 Illustration of traditional Swedish approach of product development processes

3.1.2 Different Roles in the Product Development Process.

Figure 4 The new roles of engineering

3.2 Lean Product Development

THE 13 PRINCIPLES OF LEAN PRODUCT DEVELOPMEMT SYSTEM

• Principle 1: Establish Customer-Defined Value to Separate Value-Added Activity from

• Principle 2: Front-Load the Product Development Process While There Is Maximum

• Principle 4: Utilize Rigorous Standardization to Reduce Variation, Create Flexibility and

• Principle 5: Develop a Chief Engineer System to Integrate Development from Start to

• Principle 6: Organize to Balance Functional Expertise and Cross-Functional

• Principle 7: Develop Towering Technical Competence in All Engineers. Since the

• Principle 9: Build in Learning and Continuous Improvement. The most sustainable

Figure 6 Example of trade-off curves

Figure 7 The value-adding cycle (Ward, 2007)

3.2.2 Definitions of Waste

Unsynchronized concurrent Producing ahead what is

THE THREE M:S (MORGAN & LIKER, 2006)

Muri (overburden): When a machine, process or person is pushed to work beyond it

3.2.2.2 Ward’s Definition of Product Development Waste

Waste caused by Scatter

Waste caused by Wishful thinking

DIFFERENT CATEGORIES OF WASTE IN PRODUCT DEVELOPMENT

3. Not finding necessary information easily: When it is a lack of accessibility to information

3.2.3 Set-Based Concurrent Engineering

Figure 10 Concept narrowing in Set-Based Concurrent Engineering

3.2.3.1 Framework for Set-Based Concurrent Engineering

3.2.4 Cost Awareness

Table 2 Example of prioritization by a weighted value

Cost Affect of Prioritization

Project number 1 Project number 2 Project number 3

Figure 11 Example of the impact on cost due to implementation order

3.2.5 Cadence and Synchronization

Table 3 Different kinds of cadence adapted from Reinerstein (2009).

Type of Cadence Description Result

Projects cascading into each other