Exploring strategic responses of the automotive industry during
the transition to electric mobility: a system dynamics approach
Christoph Mazur *>", Marcello Contestabile ‘, Gregory J. Offer *“, N.P. Brandon *”
*Energy Futures Laboratory, Imperial College London, Exhibition Road, London, SW7 2AZ, UK
> Earth Science and Engineering Department, Imperial College London, London, SW7 2AZ, UK
° Centre for Environmental Policy, Imperial College London, Exhibition Road, London, SW7 2AZ, UK
4 Mechanical Engineering Department, Imperial College London, London, SW7 2AZ, UK
*christoph.mazur@ imperial.ac.uk
We gratefully acknowledge funding from the Grantham Institute for Climate Change
(Imperial College London), the Climate Knowledge and Innovation Community (European
Institute of Innovation and Technology) and the UK EPSRC through the grant ’*°SUPERGEN
14: Delivery of Sustainable Hydrogen’’ and a Career Acceleration Fellowship for Gregory
Offer, award number EP/100422X /1.
Abstract
This paper outlines a model archetype that can be used to assess the effects of future policy
making and the future transition towards electric vehicles on the automotive sector, while
taking into account insights from innovation, transition literature and the multilevel
perspective. In order to show the flexibility of the model structure and tackle the gap on how
the automotive industry normally responds on those factors, the approach is then used
together with historical data to generate insights on how industry has responded to pressures
in the regime in the past. For that a case study approach is taken when a timelines for the
automotive regime and landscape are presented and then put in relation to a timeline of
BMW’s activities. While the study is in an early stage, still it is shown how first quantitative
parameters can be identified. The article concludes with an outline of future work.
Keywords
Policy making, automotive industry, transition scenarios, diffusion, economics, model
architecture
1. Introduction
This paper presents how the transition of the automotive industry towards electric road
transport can be explored with the help of a system dynamics driven approach, while drawing
on knowledge from transition science and especially the multi-level perspective approach.
While those qualitative approaches provide insights on experiences on past transitions as well
as means to structure systems that are in transition, our work has the aim to discuss the effects
of transitions in a quantitative way. For that we propose a system dynamics model structure
that is tailored towards the research problem. After a presentation of the different
components, the model structure is then taken as a basis to explore the nature of the system
and especially of the stakeholder ‘automotive industry’.
1.1 Background
Over the last years the automotive industry has been experiencing a number of different
pressures. Not only are they increasingly perceiving the first effects of oil scarcity, but, more
and more, these industries are getting aware of changing customer expectations and
behaviour and are challenged by governmental policies driven by climate change issues. As a
1
result, high fuel prices or the on-going discussion on emissions has led and will lead to a
variety of changes in behaviour, responses in strategies and products in the automotive
industry; especially as light-duty vehicles, such as automobiles are responsible for a high
amount of energy-related GHG emissions (26% in OECD) (IEA 2010a, WEC 2011). One
way of addressing those pressures is the introduction of technologies such as HEVs, PHEVs,
BEVs and FCVs (D. Howey & Martinez-Botas. 2010, IEA 2010b) as it is argued that the
whole spectrum of those electric powertrain technologies are likely to be needed in a future
decarbonised road transport system, each playing a different role (IEA 2010b, McKinsey &
Company, 2010). Scenarios, such as those analysed by the IEA and World Energy Council,
highlight futures with a fast rapid diffusion of PHEVs, BEVs and FCEVs (IEA 2010a, WEC
2011). As a result the automotive industry will be faced with consequences that are difficult
to predict, implying risks, but also opportunities (especially for new actors). They are facing
high uncertainty with regard to decisions conceming future technologies and strategies
(Bailey et al. 2010, Hellman & van Den Hoed 2007, Whitmarsh & Kohler 2010).
Due to the potentially significant changes to the current structure of the automotive industry
that a transition to electric mobility would induce, it is clear that for national governments
this becomes a question of industrial policy just as well as of energy and environmental
policy. With a transition towards electric mobility this paper means a move away from
internal combustion engine vehicles (ICEVs) towards cleaner vehicles, such as hybrid electric
(HEVs), plug-in hybrid electric (PHEVs), battery electric (BEVs) or hydrogen fuel cell
electric vehicles (FCEVs).
While energy and environmental policy goals are largely similar across European countries,
industrial policy goals can be expected to reflect the particular structure and strategy of
national industries and therefore vary more significantly. This hypothesis is supported by the
fact that recent policies aimed at promoting electrification of road transport have taken
somewhat different forms in different European countries (Elzen & Wieczorek 2005, Huétink
et al. 2010, Santos et al. 2010, Stern 2007, van den Hoed 2007).
But even then, current road transport policies seem to have failed to address those mentioned
issues in an appropriate way, especially as the uptake of electric cars has been slow (Huétink
et al. 2010, Santos et al. 2010, van den Hoed 2007).
However, the problem (or second issue) is that the specific impact and efficiency of those
measures is uncertain. So it is difficult to link specific developments with the effects of
individual policy measures, especially as those policies (and the whole transition itself) not
only affects uptakes of technologies, but also have significant impact on a whole system of
relevant stakeholders including the automotive OEMS or their suppliers. Therefore, in order
to understand the consequences and outcomes of those policies, it is necessary to understand
the transition process with its different stakeholders, their behaviours as well as the relations
between them. Only then it will be possible to assess the efficiency of measures and choose
the most efficient ones. This is also of relevance when decision makers want to feedback the
outcomes of their policies into reviews of those policies, creating feedback loops.
This knowledge is also of interest to the various industrial players, whose short-term
strategies currently seem to lock them to combustion engines, limiting their investments into
alternative technologies and waiting for anticipated spill-over's from other companies who
are executing this research (E4tech March 2007, Santos etal. 2010). Understanding the
complex system they are in as well as knowing the consequences of their own strategies and
decisions, can decrease those uncertainties, in order to reach their business objectives.
1.2 Understanding transitions of the automotive sector
There are two major ways to discuss the transition processes; qualitative studies mainly based
upon insights of innovation sciences, and quantitative studies that use modelling to explore
and to compare the various effects.
Qualitative studies such as (Bakker 2010, Collantes 2007, Farla et al. 2010, Pinkse & Kolk
2010, Santos et al. 2010, van den Hoed 2005, van den Hoed 2007, Wiesenthal et al. 2010),
for example, outline the challenges as well as relationships between the various actors, and
describe their roles and significance for the diffusion of electric mobility (Collantes &
Sperling 2008, Kieckhafer et al. 2009, Schwanen et al. 2011). For that, a variety of studies
((Nykvist & Whitmarsh 2008, Suurs et al. 2009)) describe and formalize transition systems
with the help of innovation management (mainly based upon (Geels 2002, Geels & Schot
2007, Rip & Kemp 1998)).
A number of authors point out the challenges for the industry, which is facing uncertainty
with regard to the decision concerning future technologies and strategies (Bailey et al. 2010,
Hellman & van Den Hoed 2007, Whitmarsh & Kohler 2010). Their activities are interpreted
as a response against pressures from external actors, like regulators, consumers or
competitors. Less than 5% of their R&D funding is directed towards technologies focusing on
electric power trains (Wiesenthal et al. 2010). The short-term strategies of industry lock them
into using combustion engines, avoiding extensive investments into alternative propulsion
technologies, as they anticipate and wait for spill-overs from other companies executing this
research (E4tech March 2007, Santos et al. 2010).
Quantitative approaches commonly use modelling techniques to simulate different diffusion
pathways for electric transport technologies. They asses the influence of various scenarios on
the simulated transition outcomes and derive from that recommendations for policy makers.
In system dynamics the transition process is modelled in a top down-approach, where the
different processes are modelled on an aggregated level (with the help of differential
equations). This equation based model assumes that agents are well mixed. It has been
demonstrated for complex systems, especially where feed-back loops are significant.
Aggregated equation-based models (see for rank, probit, stock approaches in (Norton & Bass
1987, Stoneman 2002), etc.) can be easily utilized in this approach. However, it is difficult to
describe interaction between individual actors, as this model operates on an aggregated basis.
Agent-based modelling, where stakeholders are represented as individual agents, is
illustrative of a bottom-up approach. A gents can be goal-directed (behaving with respect to
their utility), reactive (responding to changes in the environment) and capable of interacting
with other agents. Agent-based modelling is seen to be the most complex and elaborate
diffusion model to date (Frenzel & Grupp 2009). Both, the system based approach (e.g. in
(Keles et al. 2008, Meyer & Winebrake 2009, Struben & Sterman 2007)) as well as the agent-
based modelling approach (e.g. in (Safarzynska & van den Bergh 2010, Sullivan et al. 2009))
were already applied to simulate different transition path of sustainable technologies. In
certain cases, both approaches were combined in order to take advantage of both models
strengths (Kieckhafer et al. 2009, Kohler et al. 2009).
In summary, quantitative approaches can provide policy makers with information conceming
the consequences of policies in terms of the diffusion of certain technologies (Holtz 2011).
For instance, they describe the transition with regard to the evolution of infrastructure
policies (Huétink et al. 2010, Meyer & Winebrake 2009, Park et al. 2011, Struben & Sterman
2007). In addition, the influence of the customer on transition outcomes, and especially the
3
effects of marketing, word-to-mouth and how these are affected by policies, can be simulated
(Struben & Sterman 2007). For example, through such simulations, (Charalabidis et al. 2011,
Holtz 2011, Keles etal. 2008, Meyer & Winebrake 2009, Park etal. 2011, Sullivan et al.
2009) one can outline and compare the influence of provision of a suitable infrastructure and
the application of subsidies in order to achieve a transition towards a fuel cell based mobility.
Although in current quantitative studies, different actor types are addressed in individual
ways, manufacturers are only formalized as aggregated providers of technologies and new
vehicles. But their actual behaviour and their effect on the transition, as well as how they are
affected by the transition process itself, is not extensively discussed, although their significant
role in the whole process has been emphasized. Also, how their individual goals, such as their
market share or capacities are affected, has not been discussed further. However, in order to
describe the transition process, it is necessary to get a better understanding of all relevant
actors, and especially of the industrial actors, as they have a substantial influence on the
diffusion/transition process. Also their influence on the transition itself is of interest. Here we
present a model structure that allows these issues to be explored.
Hence, in this work we do not intend to outline an approach where future transition or
diffusion scenarios shall be the outcome, but instead present a model structure with very
narrow boundaries that focus on the behaviour of actors with respect to exogenous transition
and diffusion scenarios in order to explore how industrial players, such as the automotive
industry would react to those scenarios.
1.3 Aims, methods and structure of the paper
While transitions, and especially the effects of policies on the system and on the automotive
industry, have been discussed by transitions science in a qualitative way, it is still difficult to
attribute the consequences of certain policies or events to specific outcomes or effects. While
the qualitative methods can indeed provide some additional insights on those matters, the
complexity of the discussed socio-technical system requires, in our opinion, a quantitative
approach.
Hence, in this paper we propose a system dynamics model structure as a quantitative
approach that could be used to describe, to attribute and to assess individual consequences
(such as transition pathways and patterns) and their causes (such as policy measures and
industrial behaviour) in order to explore and understand different transition scenarios.
The choice of transition theory as a framework for the analysis is justified by its ability to
capture all key dimensions of a transition process such as the one under study, which other
disciplines are not able to do (this is further discussed in Section 2.1). A variety of works
(Geels 2005a, Ieromonachou etal. 2007, Nykvist & Whitmarsh 2008) already discuss
transitions in road transport from a historical point of view, with current research (Van Bree
etal. 2010) outlining possible transition futures and scenarios (and future technology
choices), using the Multi-level perspective (MLP) drawing upon insights from historical
transition pathways (Geels & Schot 2007).
The structure of the paper is as follows. Section 2 presents the proposed model structure with
the different domains that define its system boundaries. Section 3 presents an adapted
proposed model structure and discuss a case study based upon that model structure. Section 4
concludes with a discussion of the approach and summarizes future work.
4
2A model structure to understand the strategic responses of the
automotive industry
This section presents the proposed model structure. Before describing the model itself, it will
first introduce the theories it is drawing from: especially multi-level perspective and
transition science.
2.1 Characteristics and specification of the model
In contrast to the approaches outlined in section 1.2 the model structure that is presented in
this paper does not intend to describe the behaviour of the whole transport sector with respect
to a transitions but instead focuses on the dynamics in the automotive industry and especially
the behaviour of individual automotive stakeholders, such as regime OEMs, suppliers or
niche actors. Also our approach does not have the aim to predict future diffusion scenarios,
but instead to use those predicted futures to create scenarios whose impact on the automotive
industry will be then tested. The system dynamics based model structure that we present shall
accommodate the following aspects (that will be presented and discussed in the subsequent
sections):
= Exogenous diffusion and transition scenarios as input (2.1.1)
= Incorporation of the characteristics of potential vehicle technologies (2.1.2)
= Based upon multilevel perspective (2.1.3)
= Transition pathways and patterns (2.1.4)
= Simple model representation to facilitate interaction with automotive actors (2.1.5)
2.1.1 Exogenous diffusion and transition scenarios as input
As mentioned in section 1, there is already a variety of studies that use methods from agent-
based modelling or system dynamics to model transitions (Huétink et al. 2010, hyeong Kwon
2012, Keles etal. 2008, Meyer & Winebrake 2009, Park etal. 2011, Shafiei et al. 2012,
Shepherd et al. 2012, Struben & Sterman 2007, Sullivan et al. 2009), some of them also take
into account insights from MLP (e.g. (Papachristos 2011)). Those models explore the aspects
of infrastructure, subsidies or policy making and have in common that they aim to assess the
consequences of those events or effects on the diffusion of electric road vehicle technologies.
They try to assess the impact on future diffusion scenarios and give insights (or forecasts)
into which alternative drive train technology might be the winner in the future.
In contrast, our work aims to assess the effects of diffusion or transition scenarios on the
automotive industry, and especially its strategic behaviour towards changes in its
environment. As a result, the above mentioned diffusion scenarios act (at this stage) as
exogenous model inputs, which then then provide scenarios to test automotive actors’
behaviour. As a source for those scenarios, our approach draws on two sources. The first, in
order to calibrate the model, are historical timelines (as presented in section 3) outlining
events that are relevant for automotive industry actors. In order to explore the effects of
possible futures this work draws on widely accepted scenarios such as those outlined by the
IEA and World Energy Council, describing different timelines for market shares for PHEVs,
BEVs and FCEVs (eg. (IEA 2010a, IEA 2010b, McKinsey & Company, 2010, WEC 2011)).
However, those studies normally only outline quantitative diffusion scenarios for the different
vehicle technologies and to certain extent also for the corresponding infrastructures, whereas
in our case we are discussing transitions of the whole system. That is necessary as those
diffusion scenarios do imply more than just changes in numbers of vehicles but instead do
5
also affect the whole system. The implications for that have been discussed in innovation
sciences with the help of the multilevel perspective that will be further discussed in section
2.3
2.1.2 Incorporation of the characteristics of potential vehicle technologies
Electrification relies on a range of vehicle powertrain technologies such HEVs, PHEVs,
BEVs and FCVs. It is argued that the whole spectrum of electric powertrain technologies are
likely to be needed in a future decarbonised road transport system, each playing a different
role (IEA 2010b, McKinsey & Company, 2010). Each of those technologies has advantages
and disadvantages in terms of energy density, price or infrastructure requirements (D. Howey
& Martinez-Botas. 2010, Offer et al. 2010). As a result it is necessary to characterize the
products of the observed automotive company and its capabilities to deliver one of those
technologies, as well as their portfolio.
2.1.3 Socio-technical systems and the MLP
This section briefly introduces the multi-level perspective that will be used in our model to
structure the system. The multilevel perspective (MLP) has been already used to discuss
aspects in transport. (Whitmarsh 2012) provides a summary on its contribution to that
domain.
A transition towards electric road transport affects a number of actors such as automotive
OEMs and suppliers, providers of infrastructures (such as oil, gas and utility
companies/suppliers), and owners of the vehicles, forming a cluster of elements that is
characterized by the presence of feedback loops and path dependence. Such a cluster is called
a ‘socio-technical system’ (Geels 2005b).
The model structure that is presented in this paper is based upon the “multi-level perspective”
approach, a major strand of current innovation research, developed by Geels, Kemp, Schot
(Geels 2005b, Rip & Kemp 1998) and others in the Netherlands since the first half of the
1990s. It largely builds on evolutionary theories of technological innovation (Geels, 2002).
The MLP approach is used as a basis for research on transitions, leading to typologies of
transition pathways (Geels & Schot 2007) and pattems (De Haan & Rotmans 2011). It
describes socio-technical systems as divided into three distinct but closely related levels: the
landscape, regime and niche level.
In the case of electric mobility the landscape represents external effects such pressures due to
climate change, rising oil prices or changed perceptions towards sustainability; however, it is
generally stable and takes a long time to change (i.e.: in the order of years or decades). Niche
and regime actors experience changes in the landscape as external pressures and respond to
them accordingly. The current transport regime is defined by a set of elements such as the use
of fossil fuels and combustion vehicles and appropriate fuel and production infrastructures as
well as beliefs and habits that are consistent with those, forming together the current road
transport system. Regimes can change under certain conditions (i.e.: pressures arising from
the landscape or from niches). Such changes, where significant, go under the name of a socio-
technical transition. Though they change faster than landscapes, regimes mainly generate
incremental innovations. In contrast to that, the generation and development of radical new
innovations is often situated in the so called niches. Niches are seedbeds for change and are
nomally relatively protected market or technological domains, where new systems and
practices appear or are tested, providing a room where new networks and the exchange of
learning processes can arise Regimes can be challenged and replaced by new regimes
6
emerging from niches, especially as pressures, induced from the landscape, can open
windows of opportunity for the new regimes (Geels 2002, Geels & Schot 2007, Kemp &
Loorbach 2003, Rip & Kemp 1998, Smith et al. 2005, Tukker & Butter 2007).
For transition science the MLP has provided a mean that allowed to structure socio-technical
systems. Based upon that and the study on past transitions, transition science has identified a
number of stereotypic transition pathways and transition patterns. Section 2.1.4 provides an
overview of those.
2.1.4 Transition pathways and patterns
As mentioned in section 2.1.3, the MLP provides a way to structure the observed system into
different levels. Combining it with historical observations of transitions of real systems,
transition science (Geels & Schot 2007) has identified a variety of triggers and drivers for
those transitions as well as barriers that support the stability of regimes. (Geels & Schot
2007) provide a typology of transition pathways. The different types are based on the nature
and timing of interactions between the landscape, the niches and the regime (see section 2.1.3
for an explanation of the various levels). Additionally to the definition of four stereotypic
transition pathways, the typology also outlines the main actors involved and the types of
interactions; providing insight that can be used as a basis for policy making (i.e. what to
target). However, transitions are not limited to purely one pathway type but also a
combination of those (depending on the persistence of the pressure induced by the landscape
and the adaptability of the regime). While (Geels & Schot 2007) provide 4 widely accepted
stereotypic transition paths, recent work (De Haan & Rotmans 2011) introduces a set of
common transition patterns and system states (i.e. conditions such as pressures on the
system). Their typology is classified by the source of transition pressure (Reconstellation,
Empowerment or A daption) and for each of those a number of possible transition processes
are outlined.
While the typologies have been used to discuss transitions towards electric road transport
derived possible futures (e.g. (Van Bree et al. 2010)), in one of our past studies ((Mazur et al.
2013)) those typologies have been applied to perform an ex-ante qualitative assessment of
government policies in the area of electric mobility. While those studies discussed the
problem from a qualitative point of view, the work presented here draws upon those insights
to translate projected diffusion scenarios into transition scenarios in order to identify aspects
that are relevant for automotive actors and to describe the dynamics of those aspects. A nother
important point is that the model structure is adaptable to any type of transition pathway.
2.1.5 Simple model representation to facilitate interaction with automotive
actors
The model structure we propose in the paper was created in such a way that it allows
scenarios to be readily created that can then put forward to actors in the automotive industry,
so that they can give feedback towards the conclusions. The aim is to bring in professionals
from industry into the process at two stages: first, in the formalization of their behaviour
towards certain situations and secondly in the evaluation and discussion of the results. In the
long term, the aim is to create based upon this model structure a survey. Findings such as the
ones presented in chapter 3 shall be then put forward to stakeholders in the automotive
industry who would be asked to respond to those claims.
2.2 Archetype of system dynamics model
In this section we present the model that shall be applied to examine the effects of transitions
and especially of policy making on automotive actors, such as established regime OEMs and
suppliers or niche actors, while incorporating the characteristics and specifications outlined in
section 2.1.
Fig. 1 illustrates the general archetype of model presented here. It allows simulating the
actors’ behaviours over time. It is composed of 5 main components. The variable AUTOR
CHARACTERISTICS represents the current state of the studied actor (automotive player). A
characteristic can be an aspect such as the existence of knowledge in Fuel Cell technology or
a division working on fuel cells, or whether the company already offers PHEVs or not, or
indicators such as the average fleet emissions. On the other side a set of REGIME
CHARACTERISTICS has to be defined that correspond to the ACTOR CHARCTERISTICS
and vice versa. At this point, the insights from transition theory come into play, as they allow
deriving those corresponding and relevant characteristics from projected diffusion pathways
(as provided by IEA). The regime and landscape characteristics are then compared with the
actor’s characteristics in order to determine DISCREPANCIES.
ACTOR REGIME
x iow a
DISCREPANCY
Induced EFFECT Detween the regime and
actor
ACTOR REACTION
executed with respect to
discrepancy
Fig. 1: General archetype of proposed model
Fig. 2: General archetype of proposed model
These are then used as an input to determine actor responses (ACTOR REACTION) to those
pressures. The reactions or decisions provide favoured goals for ACTOR
CHARACTERISTICS (e.g. the decision to provide a PHEV). The induced EFFECT captures
the transient behaviour of the system. This can be, for example, the time it takes until a
production capacity is created, or the time it takes to develop a first demonstrator. So this can
be parameters derived from historic data, as well as leaming curves or economies of scales,
depending on the nature of the variable and goals.
As one can see the definition of the relevant ACTOR CHARACTERISTICS and REGIME
CHARACTERISTICS is crucial. Here insights from transition science help defining the
system boundaries. In the term ACTORS REACTION it is necessary to determine, how
actors react towards certain discrepancies. For that, insights from literature, surveys, as well
as the application of the model itself, can provide inputs (see section 3).
In comparison to Fig. 1, Fig. 3 illustrates a more problem specific model. In this step the
influence of expectations on the regime and landscape are taken into account, as they play a
crucial role in the decision process of the automotive industry (Bakker et al. 2012, Budde
8
et al. 2012). Furthermore, this model architecture differentiates between the PRESSURES the
different DISCREPANCIES induce. This helps answer the question; do automotive actors
respond to different pressure and incentives in different ways, depending on their own
CHARACTERISTICS and options (e.g. currently available resources)?
CURRENT REGIME
SOCIO-TECHNICAL
ACTOR ATTRIBUTE that CHARACTERISTIC (quantitative)
corresponds with attribut of
socio-technical regime (quantitative)
EXPECTED REGIME
SOCIO-TECHNICAL
DIRECT EFFECT on acto! CHARACTERISTIC tative
anche ant imei DISCREPANCY between the (quantitative)
juantitative) regime and actor
ia (quantitative/qualitative)
ACTOR REACTION executed By Actor perceived
with respect to discrepancy and PRESSURE
pressure (qualitative/quantitative) (qualitative/quantitative)
Fig. 3: Detailed archetype of proposed model
It can be seen that the model presented here is a simple illustration of the decision process of
automotive actor facing changes in the regime or landscape, and who has to respond towards
those pressures. Not only is it a simple decomposition of the whole problem and focuses on
the crucial aspects of the system, but also a simple illustration of the problem, making it
usable for the communication and interaction with actors (e.g. interviews).
There is no feedback from the actor towards the regime itself at this stage (for instance,
influence on market due to success of a product, or possible lobbying).
3 Understanding the automotive industry: the case of BMW
Section 2 presents the model structure that was used to explore the effects of different
transition scenarios as well as policy making on the automotive industry. To illustrate the
flexibility of the model structure a study with the aim to generate model parameters was
carried out. As mentioned before the behaviour of the automotive industry is not well
understood. In order to fill that gap, we have adapted the model structure outlined in section
2, so that it can be applied to understand that actors’ behaviour (see Fig. 4 for the adapted
model).
First the adapted model will be presented in section 3.1. This will be followed by the
description of a study on BMW's behaviour during the last 20 years, based upon the model.
3.1 Methodology
The model presented in section 2 has the aim to assess the effects of policy making and future
transitions on industrial actors, taking into account future scenarios. In such a case different
projected timelines (by IEA and others) represent model inputs to explore how the system
reacts to those inputs. However, in order to execute such a simulation the other various
components have to be specified first. While the effects within the automotive industry (actor
attributes) can be derived from literature it is difficult to retrieve objective insights on how
decisions are taken within the industry and what pressures or goals play a dominant role.
CURRENT REGIME
SOCIO- TECHNICAL
ACTOR ATTRIBUTE that CHARACTERISTIC (quantitative)
corresponds with attribut of
socio-technical regime (quantitative)
EXPECTED REGIME
SOCIO-TECHNICAL
DIRECT BREET onacior DISCREPANCY between the CHARACTERISTIC (quantitative)
ruantitative regime and actor
fa a) (quantitative/qualitative)
ACTOR REACTION executed By Actor perceived
with respect to discrepancy and PRESSURE
pressure
Fig. 4: Adapted model for application in case study
In order to obtain that data for our future work, historical insights are used as a basis for the
determination of that behaviour. So instead of using projected data for the regime and
landscape variables, this study applies available historical data in order to understand the past
behaviour of automotive actors. For that, the loop was interrupted at the position where the
behaviour of the actor is normally described. Then information on the landscape, the regime
and the automotive actors (in this case BMW) is extracted from the literature, journals or
BMW's annual reports, and timelines for both sides are created: BMW and its action on the
one side, and what was happening around BMW on the other. Those are then aligned to
obtain (at this stage) qualitative insights on how BMW responded towards the various events
and pressures and how the companies’ characteristics (such as knowledge or products) were
affecting those.
10
While the study is of qualitative nature at this point, it still represents an important step on the
way towards the quantification of the whole system and its dynamics.
3.2 The case of BMW
With respect to the model that has been presented in section 2, this section will in brief
outline how the landscape and regime have developed over the last 20 years, and provide
relevant milestones that BMW has experienced during that period. Both are then put in
relation to each other and based upon a brief comparison, a set of insights are derived. There
are a variety of dimensions that describe the socio-technical system for road transport. This
study only concentrates on the major aspects that have been identified to have major impact
on strategies in the automotive industry. Extracted from a number of studies (Bakker &
Budde 2012, Bakker et al. 2012, Collantes & Sperling 2008, Dijk & Yarime 2010, Hacker
et al. 2009, IEA 2011, Kohler et al. 2012, Mazur et al. 2013) those aspects include technology
trends and hypes, national and international policies, BMW’s competitors, economic
pressures, fuel prices and infrastructures. As expectations play an important role I the
automotive industry (Budde et al. 2012, Konrad et al. 2012), the timeline shall also give what
mood was perceived during the respective period of time. The summaries and timelines that
are presented here are mainly based upon these papers, and are additionally supplemented
with actual data on economic growth or fuel prices as well as the perception of the media
(mainly drawing from articles of major German (car) journals and newspapers such as the
Spiegel, Focus, Autobild or Handelsblatt and Frankfurter Allgemeine Zeitung with (Auto
Bild 2002, BMW GROUP 2003, Der Spiegel 1996, Der Spiegel 2005, Der Spiegel 2006,
FOCUS 2009, Frankfurter Allgemeine Zeitung 2001, Frankfurter Allgemeine Zeitung 2005a,
Frankfurter Allgemeine Zeitung 2005b, Frankfurter Allgemeine Zeitung 2008, Handelsblatt
2009, Spiegel Online 2003, Spiegel Online 2006, Spiegel Online 2011, VDI Nachrichten
2010) as few important articles, mainly around the International Automobile Exhibition IAA
in Frankfurt or Detroit Motor Show, that outline the mood of the whole industry and the
trends and fashions at those times. The information on BMW is additionally supplemented by
an analysis of the annual reports of the BMW Group. The following two summaries (of the
landscape/regime and BMW) are based upon the sources outlined here until now. The main
events and aspects are outlined in the figure Fig. 5.
3.2.1 The regime and landscape from the viewpoint of the automotive
sector (1990 - 2012)
The Zero Emissions Vehicle (ZEV) initiative that outlined a number of emission targets and
limits as well as diffusion goals for zero emission vehicles in California in 1990 was one of
the first incentives pushing towards low carbon transport. Although limited to California, it
had a huge signalling effect as around 25% of the US vehicle market was in California and
developments in Califommia were expected to move to further states. However, though it
triggered a number of EV and FCEV prototypes being presented by the industry, the fact that
those goals were not expected to be met meant it was relaxed in 1996, and a key regulative
pressure on the automotive industry was relaxed. While until then hydrogen vehicles had
been seen as a solution for future transport, suddenly the mood changed and interest switched
towards other technologies such as to the development of battery electric drive trains. Around
1996/97, at a time where hydrogen was not seen as a winner anymore, major OEMs in
Germany and Japan presented their respective solutions to deal with arising discussions on
CO, emissions. While Daimler’s launch of the NecarlII hydrogen demonstrator surprisingly
triggered a new hydrogen/fuel cell hype (especially in Germany), Toyota’s and Honda’s
launch of hybrid electric vehicles (Prius and Insight) directly on the Japanese and USA
11
markets was acknowledged by the automotive sector without having initially any disruptive
effects. While the early 2000s were dominated by an economic world crisis it was not until
2004/05 that major changes were triggered in the automotive sector. The success of Toyota
Prius HEV and the launch of its second generation, as well as rising fuel prices lead, to a
Change in the perception of the hybrid technology, that until now lead to a ‘hybrid race’
illustrated by the significant increase in hybrid patents, HEV/PHEV prototypes being
presented at various automobile exhibitions as well as numerous announcements of HEV
release dates. During that time, hydrogen technology research, though not promoted
anymore, still enjoyed support by various governments (like for example the US Department
of Energy Hydrogen Program). However, with the inauguration of Steven Chu as new US
Secretary under President Obama in 2009 and a reassessment of all technology options, the
perception and expectations conceming that technology completely changed. Only the
intervention of the Congress and the Hydrogen lobby could stop USA hydrogen R&D funds
being completely cancelled. During that time (a time where the financial crisis hit)
governments such as the German (Nationale Platform Elektromobilitat) or the UK (Ultra Low
Emission Vehicles initiative) launched national programs supporting the uptake of
electromobility in order to reach environmental or industrial target. Since then, HEV/PHEVs
and EVs have dominated current discussions (hydrogen no longer exists in the US White
House Blueprint for Secure Energy Future), and the presence of TESLA in the media, the
introduction of many EV demonstrator projects (SmartEV, MiniE, and many more) as well as
the recent introduction of the Chevrolet Volt, Nissan Leaf or Mitsubishi iMiEV support that
impression. While the financial crisis as well as rising fuel prices in the late 2000s put the
focus on the development of small and highly efficient vehicles, the current (2012/13) focus
has switched again slowly but steadily towards SUVs and PHEVs and EVs figure in
technology portfolios as short- or medium-term solutions.
3.2.1 BMW (1990 - 2012)
BMW is one of the major German automotive car manufacturers with more than 1.6 million
cars sold, a profit of more than € 7 billion and more than 100,000 employees (2011). During
the last 20 years BMW has increased its output in vehicles from less than 900,000 and
increased profits nearly tenfold. Since then BMW has had some experience with hydrogen
vehicles (both combustion and FCEV), battery electric Minis and is currently launching its
first hybrid vehicles, and especially its i series (BMW i3). While it has been always seen as
the smaller premium automobile manufacturer, its sales numbers overtook those of Daimler a
few years ago. BMW’s image is located around medium/large luxury and performance
segment cars, a fact that is well reflected in the average fleet emissions.
While in the early 1990s ZEV regulation driven experiences with alternative vehicle power
technologies were disappointing, it was not until 1996 that BMW established serious
hydrogen research activities. Though it was in a time of hydrogen disappointment in the
automotive sector, the decision was mainly motivated by a hydrogen vehicle demonstration
by its main competitor, Daimler. Though the work focused on PEM fuel cells and later SOFC
fuel cells as well, BMW presented in 1998 the BMW 750hL, a large segment vehicle with a
hydrogen combustion engine and a 5kW PEM APU. Since then, BMW built more than 100
hydrogen combustion vehicles that were used at various events (such as the EXPO 2000 in
Germany) and a number of demonstrator programs where those vehicles proved themselves
in over more than 4,000,000 kilometres driven. Also, a petrol fuelled vehicle using a SOFC
APU with a fuel processor instead of an alternator was introduced. Though BMW announced
it would bring its hydrogen combustion vehicle to market in 2002, those vehicles did not go
beyond the status of demonstrator fleet programs.
12
While the average fleet emissions of BMW had heen stable (though at a relatively a high
level), emission target discussions at the European level had led in the late 90’s and the
beginning of the 2000’s to an introduction of a variety of engine efficiency improvements as
well as the higher use of diesel in the fleet, leading to a slow but steady decrease in average
fleet emissions. However, hybrid or electric vehicle development was not intensified in this
period of time. The acquisition of Rover brought in Range Rover SUV technology, leading to
the design and production of BMW’s X5 SUVs. Also, the early 2000’s were more focused on
the world economic crisis, the new emerging Chinese market and the opening of BMW’s new
production facility in Leipzig in 2005. This changed in 2005/06, mainly with the success of
Toyota’s Prius and rising fuel prices, causing customers to demand similar solutions. Had
until then, only the hydrogen technology featured in the annual reports as a future solution for
low emission vehicles, from 2005/06 on, the hybrid vehicle technology had been included in
the annual reports as well. Around that time a collaboration with GM and Daimler-Chrysler
was announced in order to develop a hybrid system to compete with the Japanese
manufacturers. Also, in 2006/2007 BMW intensified its hydrogen combustion vehicle
activities by leasing out 100 vehicles. In this period, BMW’s average fleet emissions were
increasingly dropping (approx. 170 gCO2/km), however, discussions on regulations in the
European Union on a limit of 130gCO2/km increasingly created pressure. Apart from
recuperating energy to its lead battery with the help of the alternator, BMW had not provided
any hybrid vehicle solution so far.
In 2007, a successful year for BMW, with no signs of the financial crisis yet to come, BMW
initiated a project called ‘project i? (under the ‘Number ONE strategy) that reviewed the
technology future options. This led to a significant change in the technology strategy of
BMW. Shortly after the review had finished, BMW stopped its combustion hydrogen vehicle
program and announced the launch of a Mini EV trial fleet, a battery collaboration with SB
LiMotive and the creation of a Joint Venture with PSA (Peugeot/Citroen). In 2010, it also
announced plans to develop and produce a BEV for the mass market, while also a hydrogen
and petrol hybrid vehicle prototype (using a 5kW PEM Fuel Cell, see above for APU) was
presented drawing upon the experiences available in house.
Today, in the early 2010’s, after a number of competitors have brought their PHEVs or BEVs
to market, BMW has presented its Megacity Vehicle (BMW i3), a small lightweight BEV
vehicle that is expected to reach the mass market in the end of 2013 (built in Leipzig). During
that phase the acquisition of SGL Carbon, the supplier of lightweight materials for the i3 was
announced. 2012/2013, in a time, where the amount of HEVs/PHEVs in BMW’s portfolio is
limited, BMW and Toyota have agreed to collaborate on fuel-cell systems, lithium-air
batteries, lightweight technologies and electric powertrains.
Currently BMW is selling more vehicles than ever and has the highest profits it has ever
made.
13
Relevant landscape & regime
for automotive industry pains
Gemany decroases funding for FC from 25 to 14m
1995 2000
{US Energy Policy Act 2005 & DOE Hydrogen
program lead increases hydrogen suppor from
$150 (2004) uo 276 mil in 2008)
TUS White House
EU discusses
20,
‘Economic eis leads to wreckage grant
National fi etictes below
ont 12002
Elektromobiuut || "71S fee
Suatogic
sll EVs by 2000
Patrol lnfastctre dominant while oes are neglecable
US Seerotry Chu amounces inp
or Fc funding congress stops tha
Dow Jones Index
Aw
Tigh increase im EV hyo patent applications
[Vehicle sates drop dv wo exis]
[royot Japa
launches 2s
gen Pris
Vehicle sas drop du to crs
High ol prices
effal
Pel ol Tor are
Diesel (ae
price
Dave to smaller cas
Bisng OF || Media denies
pices Soon as |] China as possible
ig probe | q|
lea |
[ Tad Tard to surzes a Toyo Ps
dectic vices
SUV boom
Technology
and vehicles
Collaboration
Ce
oa <a a Ta
can Ee Eee Boo Pa
BMW ceo cas | : ee Cee ee
oo To angen or Nag ala
et Eee
Timer ‘amen coy 7 ToT [eT
ae Ew eo
Ning acl Rolls Royce aid Colaberetion, —s#9#00 st bd ils uruad
‘conceming future ‘Hydrogen with Daimler & GM. for 2009 ann Focus on hybrid BMW 13 for 2013
= am pees ee Peon pee
wie | fdtpueke | | “sit See frees
7 A
: 2 we ata
isappointing experiences wit ‘and Pests with incremental efficiency improvement Change of opinion a ‘Scheduled for
j Peoeeting avant Fe ea common ll ised vlveuone gel mom projet | | “Gemany 2014
Getee Fen] a, toon] [150m oem [ tvmeonen | P| TOR ve TS
oar ae ed Sor Ee ae somtee | ame,
: aoe Lames Se aa ne Se _
Steg So Panis
area Es
Sirs
Tome
oe ph ER
eat conf =~ oir |
q--p=4 7
HES ee 2
Sc ee Shs, 200
2000 | ~~|-. - 0h 20
6000) | y s 120
Sa ——- 400} 170
‘at ell SE SES et Sa SS 160
2000 ni x =e 5 =P pe
1990 1995p) 4 2000 2005 NESS] 2010 es
Exemplary timeline of the regime and landscape relevant for an automotive OEM and BMW
14
3.3 Results
Based upon the system thinking outlined in Fig. 4, this section outlines the different relations
that have been identified in a first study of the data provided in Fig. 5. Also it is outline how
those results can be translated and used in a system dynamics model.
Two types of insights on the behaviour of this respective automotive OEM could be made
based upon this case study: what are the factors that are taken into account for the choice of
the direction, and when is a decision process triggered.
As an example, in terms of BMW it seems that the choice of power train technology was not
affected by government research subsidies. Moreover, it chose technologies that were
currently popular as well as suiting their current knowledge and product portfolio. However it
did not change towars a popular technology in the moment it had become popular. Instead,
triggers such as the launch of a product by a major competitoir (e.g. Necar II by Daimler) or
the high pressure from the public and the success of a competitor (such as the 2nd generation
Prius) were needed to trigger a change or a decision process, or just the lauch of a review
project (project i). Also, oil prices or sales crisis had small influence on the research and
product strategy.
So in this case the approach we have taken in this work can offer the indicators that are
affecting decisions, and also what the triggers are for those decisions.
Furthermore such a study also offers insights on timelines and durations. As it can be seen by
the BMW X65 project, or setting up the hydrogen PEMFC and SOFC programmes, there are
common patterns in how long such processes take. Those durations can be then implemented
into the future model. BMW founded a PEMFC team in 1997 and the SOFC project in 1999
and in each case they took 2 years to build a prototype APU. The existence of a division or
expertise in a certain technology domain can be speciified with the help of boolean variables
and then taken into account in the decision process simulation. Fig. 6 illustrates an example
for the illustration of an actor.
Those are just two short examples of the type of information that can be extracted and the
type of quantification can be achieved. This information can be used to parameterize the
decision process in order to test the behaviour of the actor with respect to future transition
scenarios.
15
1990 1995
2000
2010
‘Kaowledge in intemal combustion engines (ICE)
Knowledge in highly efficient combustion engines (ICE)
Fist apeiaae wih loon | Tenth experiences wih fol el (FC)
Expattise
BMW characteristics
Fist expatiences with hybrid power tains
segment lage cars
Products
segments
E-sogment executive cars
Fesegment cary cas
Range Rover.
‘Fsegment sport uility cars (SUV)
Fig. 6: Characterization of observed actor
4 Conclusion and future work
In this paper we have presented a system dynamics model archetype and approach that can be
used to explore the effects of policy making and transition scenarios on actors in the
automotive industry, while using insights from a variety of research domains such as
transition science or the multi-level perspective. In contrast to past works where system
dynamics has been used to outline the effects of certain policies or events on the diffusion of
vehicle technologies, this approach has the aim to use those scenarios as input parameters in
an exogenous way and to test their influence on the future behaviour of the observed actors.
However, as data on the behaviour of the automotive industry is limited, the model approach
is here adapted and used to show how data from the past can be used to obtain insights on
behaviour, and especially decision making. For that a case study looking the past 20 years of
BMW and the relevant automotive regime has been presented.
Based upon that a set of relations is outlined, showing how approach presented here can be
used to obtain insights and model parameters that can be then utilized in a model to assess
future scenarios.
Our future work now concentrates on the analysis of a number of automotive actors in order
to understand their behaviour and to be able to derive parameters that then allow a
quantitative discussion of the effects of future transitions on the industry.
17
References
Auto Bild (2002), ‘Wasserstoff-bmw geht in serie’. http://www.autobild.de/artikel/das-quot-
0-liter-auto-quot-36299.html
Bailey, D., de Ruyter, A., Michie, J. & Tyler, P. (2010), ‘Global restructuring and the auto
industry’, Cambridge Journal of Regions, Economy and Society 3(3), 311.
Bakker, S. (2010), ‘The car industry and the blow-out of the hydrogen hype’, Energy Policy
38(11), 6540-6544.
Bakker, S. & Budde, B. (2012), ‘Technological hype and disappointment: lessons from the
hydrogen and fuel cell case’, Technology Analysis & Strategic Management 24(6), 549-563.
Bakker, S., van Lente, H. & Meeus, M. T. (2012), ‘Credible expectations—the us department
of energy’s hydrogen program as enactor and selector of hydrogen technologies’,
Technological Forecasting and Social Change .
BMW GROUP (2003), “Mehr leistung bei weniger verbrauch’.
http://www.bmwgroup.com/d/0_0_www_bmwgroup_com/forschung_entwicklung/science_cl
ub/veroeffentlichte_artikel/2003/news200322.html
Budde, B., Alkemade, F. & Weber, K. M. (2012), ‘Expectations as a key to understanding
actor strategies in the field of fuel cell and hydrogen vehicles’, Technological Forecasting
and Social Change .
Charalabidis, Y., Loukis, E. & Androutsopoulou, A. (2011), Enhancing participative policy
making through modelling and simulation: A state of the art review, in ‘European,
Mediterranean & Middle Eastern Conference on Information Systems’.
Collantes, G. (2007), ‘Incorporating stakeholders’ perspectives into models of new
technology diffusion: The case of fuel-cell vehicles’, Technological Forecasting and Social
Change 74(3), 267-280.
Collantes, G. & Sperling, D. (2008), ‘The origin of california’s zero emission vehicle
mandate’, Transportation Research Part A: Policy and Practice 42(10), 1302-1313.
D. Howey, R.N. & Martinez-Botas., R. (2010), ‘Road transport technology and climate
change mitigation’, Grantham Institute for Climate Change, Imperial College London.
De Haan, J. & Rotmans, J. (2011), ‘Patterns in transitions: Understanding complex chains of
change’, Technological Forecasting and Social Change 78(1), 90-102.
Der Spiegel (1996), ‘Energie - vision auf dem abstellgleis (der spiegel 45/1996)’.
http://www.spiegel.de/spiegel/print/d-9114902.html
Der Spiegel (2005), ‘Automobile wettlauf ums doppelherz (der spiegel 37/2005)’.
http://www.spiegel.de/spiegel/print/d-41768234.html
Der Spiegel (2006), ‘Automobile griiner schluckspecht (der spiegel 46/2006)’.
http://www .spiegel.de/spiegel/print/d-49533683.html
18
Dijk, M. & Yarime, M. (2010), ‘The emergence of hybrid-electric cars: Innovation path
creation through co-evolution of supply and demand’, Technological Forecasting and Social
Change 77(8), 1371-1390.
E4tech (March 2007), A review of the UK Innovation System for Low Carbon Road
Transport Technologies, chapter 3 General Features of Innovation Systems fopr Low Carbon
Transport Technologies, pp. 15-28.
Elzen, B. & Wieczorek, A. (2005), ‘Transitions towards sustainability through system
innovation’, Technological forecasting and social change 72(6), 651-661.
Farla, J., Alkemade, F. & Suurs, R. (2010), ‘Analysis of barriers in the transition toward
sustainable mobility in the netherlands’, Technological Forecasting and Social Change
77(8), 1260-1269.
FOCUS (2009), “Bmw: Aus fiir den wasserstoffantrieb’.
http://www. focus.de/auto/news/bmw-aus-fuer-den-wasserstoffantrieb_aid_460849.html
Frankfurter Allgemeine Zeitung (2001), ‘Alternative energien der lange weg zu
brennstoffzelle und wasserstoff’. http://www. faz.net/aktuell/gesellschaft/alternative-energien-
der-lange-weg-zu-brennstoffzelle-und-wasserstoff-128902.html
Frankfurter Allgemeine Zeitung (2005a), ‘Autoimport - jetzt kommen die chinesen’.
http://www. faz.net/aktuell/wirtschaft/autoimport-jetzt-kommen-die-chinesen-1257479.html
Frankfurter Allgemeine Zeitung (2005b), ‘Faz.net-spezial die hybrid-welle rollt’.
http://www. faz.net/aktuell/wirtschaft/untermehmen/faz-net- spezial-die-hybrid-welle-rollt-
1256917.html
Frankfurter A Ilgemeine Zeitung (2008), ‘Automobilindustrie- es gab schon schénere zeiten’.
http://www. faz.net/aktuell/wirtschaft/unternehmen/automobilindustrie-es-gab-schon-
schoenere-zeiten-1514788.html
Frenzel, A. & Grupp, H. (2009), ‘Using models of innovation diffusion to forecast market
success: a practitioners’ guide’, Research Evaluation 18(1), 39-50.
Geels, D. (2005a), ‘The dynamics of transitions in socio-technical systems: a multi-level
analysis of the transition pathway from horse-drawn carriages to automobiles (1860-1930)’,
Technology Analysis & Strategic Management 17(4), 445-476.
Geels, F. (2002), ‘Technological transitions as evolutionary reconfiguration processes: a
multi-level perspective and a case-study’, Research policy 31(8-9), 1257-1274.
Geels, F. (2005b), Technological transitions and system innovations: a co-evolutionary and
socio-technical analysis, Edward Elgar Publishing.
Geels, F. & Schot, J. (2007), ‘Typology of sociotechnical transition pathways’, Research
policy 36(3), 399-417.
Hacker, F., Harthan, R., Matthes, F. & Zimmer, W. (2009), ‘Environmental impacts and
impact on the electricity market of a large scale introduction of electric cars in europe-critical
review of literature’, Report of the European Topic Centre on Air and Climate Change .
19
Handelsblatt (2009), ‘Auto-visionen elektrostudien und ps-boliden’.
http://www. handelsblatt.com/video/auto-visionen/elektrostudien-und-ps-
boliden/3261610.html
Hellman, H. & van Den Hoed, R. (2007), ‘Characterising fuel cell technology: Challenges of
the commercialisation process’, International J ournal of Hydrogen Energy 32(3), 305-315.
Holtz, G. (2011), ‘Modelling transitions: An appraisal of experiences and suggestions for
research’, Environmental Innovation and Societal Transitions 1(2),167 -— 186.
http://www.sciencedirect.com/science/article/pii/S2210422411000335
Huétink, F., der Vooren, A. & Alkemade, F. (2010), ‘Initial infrastructure development
strategies for the transition to sustainable mobility’, Technological Forecasting and Social
Change 77(8), 1270-1281.
hyeong Kwon, T. (2012), ‘Strategic niche management of alternative fuel vehicles: A system
dynamics model of the policy effect’, Technological Forecasting and Social Change
79(9), 1672 — 1680. http://www.sciencedirect.com/science/article/pii/S004016251200145X
IEA (2010a), Energy Technology Perspectives 2010: Scenarios & Strategies to 2050,
International Energy Agency.
IEA (2010b), ‘Energy technology perspectives 2010: Scenarios and strategies to 2050’.
IEA (2011), ‘Hybrid and electric vehicles: The electric drive captures the imagination’.
Ieromonachou, P., Potter, S. & Warren, J. (2007), ‘A strategic niche analysis of urban road
pricing in the uk and norway’, European Journal of Transport and Infrastructure Research
7(1), 15-38.
Keles, D., Wietschel, M., Most, D. & Rentz, O. (2008), ‘Market penetration of fuel cell
vehicles-analysis based on agent behaviour’, International journal of hydrogen energy
33(16), 4444-4455.
Kemp, R. & Loorbach, D. (2003), Govemance for sustainability through transition
management, in ‘Open Meeting of Human Dimensions of Global Environmental Change
Research Community, Montreal, Canada’, Citeseer.
Kohler, J., Schade, W., Leduc, G., Wiesenthal, T., Schade, B. & Espinoza, L. T. (2012),
‘Leaving fossil fuels behind? an innovation system analysis of low carbon cars’, Journal of
Cleaner Production (0), —
http://www.sciencedirect.com/science/article/pii/S0959652612005136
Kieckhafer, K., Walther, G., Axmann, J. & Spengler, T. (2009), Integrating agent-based
simulation and system dynamics to support product strategy decisions in the automotive
industry, in ‘Winter Simulation Conference (WSC), Proceedings of the 2009’, IEEE,
pp. 1433-1443.
Kohler, J., Whitmarsh, L., Nykvist, B., Schilperoord, M., Bergman, N. & Haxeltine, A.
(2009), ‘A transitions model for sustainable mobility’, Ecological economics 68(12), 2985—
2995.
Konrad, K., Markard, J., Ruef, A. & Truffer, B. (2012), ‘Strategic responses to fuel cell hype
and disappointment’, Technological Forecasting and Social Change .
20
Mazur, C., Contestabile, M., Offer, G. & Brandon, N. (2013), Policy making in the uk and
germany: on a transition pathway towards electric mobility? in submission.
McKinsey & Company, (2010), ‘A portfolio of power-trains for europe: a fact-based analysis
- the role of battery electric vehicles, plug-in hybrids and fuel cell electric vehicles.’.
Meyer, P. & Winebrake, J. (2009), ‘Modeling technology diffusion of complementary goods:
The case of hydrogen vehicles and refueling infrastructure’, Technovation 29(2), 77-91.
Norton, J. & Bass, F. (1987), ‘A diffusion theory model of adoption and substitution for
successive generations of high-technology products’, Management Science pp. 1069-1086.
Nykvist, B. & Whitmarsh, L. (2008), ‘A multi-level analysis of sustainable mobility
transitions: Niche development in the uk and sweden’, Technological Forecasting and Social
Change 75(9), 1373-1387.
Offer, G., Howey, D., Contestabile, M., Clague, R. & Brandon, N. (2010), ‘Comparative
analysis of battery electric, hydrogen fuel cell and hybrid vehicles in a future sustainable road
transport system’, Energy Policy 38(1), 24-29.
Papachristos, G. (2011), ‘A system dynamics model of socio-technical regime transitions’,
Environmental Innovation and Societal Transitions 1(2),202 -— 233.
http://www.sciencedirect.com/science/article/pii/S2210422411000347
Park, S., Kim, J. & Lee, D. (2011), ‘Development of a market penetration forecasting model
for hydrogen fuel cell vehicles considering infrastructure and cost reduction effects’, Energy
Policy.
Pinkse, J. & Kolk, A. (2010), ‘Challenges and trade-offs in corporate innovation for climate
change’, Business Strategy and the Environment 19(4), 261-272.
Rip, A. & Kemp, R. (1998), ‘Technological change. in: Rayner s., malone el (editors).’,
Human Choice and Climate Change. Vol. II, Resources and Technology pp. 327-399.
Safarzynska, K. & van den Bergh, J. (2010), ‘Evolutionary models in economics: a survey of
methods and building blocks’, Journal of Evolutionary Economics 20(3), 329-373.
Santos, G., Behrendt, H. & Teytelboym, A. (2010), ‘Part ii: Policy instruments for
sustainable road transport’, Research in Transportation Economics 28(1), 46-91.
Schwanen, T., Banister, D. & Anable, J. (2011), ‘Scientific research about climate change
mitigation in transport: A critical review’, Transportation Research Part A: Policy and
Practice .
Shafiei, E., Stefansson, H., Asgeirsson, E.1., Davidsdottir, B. & Raberto, M. (2012),
‘Integrated agent-based and system dynamics modelling for simulation of sustainable
mobility’, Transport Reviews (ahead-of-print), 1-27.
Shepherd, S., Bonsall, P. & Harrison, G. (2012), ‘Factors affecting future demand for electric
vehicles: A model based study’, Transport Policy 20, 62-74.
Smith, A., Stirling, A. & Berkhout, F. (2005), ‘The governance of sustainable socio-technical
transitions’, Research Policy 34(10), 1491-1510.
21
Spiegel Online (2003), ‘Suv-boom: Hochridrig durch den groSstadtdschungel’.
http://www.spiegel.de/auto/aktuell/suv-boom-hochraedrig-durch-den-grossstadtdschungel-a-
242399. html
Spiegel Online (2006), ‘Bmw hydrogen 7: Mit zisch in die zukunft’.
http://www. spiegel.de/auto/aktuell/bmw-hydrogen-7-mit-zisch-in-die-zukunft-a-436588.html
Spiegel Online (2011), ‘Iaa-auftakt: Spritfresser, die stars der show’.
http://www .spiegel.de/auto/aktuell/iaa- auftakt-spritfresser- die-stars-der-show-a-785878.html
Stern, N. (2007), The economics of climate change: the Stern review, Cambridge Univ Pr.
Stoneman, P. (2002), The economics of technological diffusion, Wiley-Blackwell.
Struben, J. & Sterman, J. (2007), ‘Transition challenges for alternative fuel vehicle and
transportation systems’, Environment and Planning B: Planning and Design, Vol. 35, No. 6,
pp. 1070-1097, 2008, MIT Sloan Research Paper No. 4587-06.
Sullivan, J., Salmeen, I. & Simon, C. (2009), ‘Phev marketplace penetration: An agent based
simulation’.
Suurs, R., Hekkert, M. & Smits, R. (2009), ‘Understanding the build-up of a technological
innovation system around hydrogen and fuel cell technologies’, International Journal of
Hydrogen Energy 34(24), 9639-9654.
Tukker, A. & Butter, M. (2007), ‘Governance of sustainable transitions: about the 4 (0) ways
to change the world’, Journal of Cleaner Production 15(1), 94-103.
Van Bree, B., Verbong, G. & Kramer, G. (2010), ‘A multi-level perspective on the
introduction of hydrogen and battery-electric vehicles’, Technological Forecasting and
Social Change 77(4), 529-540.
van den Hoed, R. (2005), ‘Commitment to fuel cell technology?:: How to interpret
carmakers’ efforts in this radical technology’, J ournal of power sources 141(2), 265-271.
van den Hoed, R. (2007), ‘Sources of radical technological innovation: the emergence of fuel
cell technology in the automotive industry’, Journal of Cleaner Production 15(11-12), 1014-
1021.
VDI Nachrichten (2010), ‘Bmw-hybrid mit brennstoffzellen’. —_ http://www.vdi-
nachrichten.com/artikel/B MW -Hybrid-mit-Brennstoffzellen/47186/2
WEC (2011), Global Transport Scenarios 2050, World Energy Council.
Whitmarsh, L. (2012), ‘How useful is the multi-level perspective for transport and
sustainability research?’, J ournal of Transport Geography 24(0), 483 — 487. <ce:title>Special
Section on Theoretical Perspectives on Climate Change Mitigation in Transport</ce:title>.
http://www.sciencedirect.com/science/article/pii/S0966692312000270
Whitmarsh, L. & KGhler, J. (2010), “Climate change and cars in the eu: the roles of auto
firms, consumers, and policy in responding to global environmental change’, Cambridge
Journal of Regions, Economy and Society 3(3), 427.
22
Wiesenthal, T., Leduc, G., Kohler, J., Schade, W. & Schade, B. (2010), ‘Research of the eu
automotive industry into low-carbon vehicles and the role of public intervention’, IPTS, JRC,
European Commission, Spain Fraunhofer Institute for Systems and Innovation Research ISI,
Karlsruhe, Germany, European Communities .
23
