Nijmegen School of Management
Nijmegen, the Netherlands
Dynamic Adaptive Policymaking for
Implementing Mobility As a Service
Dr tech P. Jittrapirom
Prof Dr Ir V.A.W.J. Marchau
Prof Dr Ir R.E.C.M. van der Heijden
Prof Dr H. Meurs
Number: Working Paper SCRIPTS-WP-18-08
Title: Dynamic Adaptive Policymaking for Implementing Mobility As a
Service (MaaS) (August, 2018)
Key words: Smart mobility; Long-term planning; Deep uncertainty; Transport
Dr tech P. Jittrapirom
Prof Dr Ir V.A.W.J. Marchau
Prof Dr Ir R.E.C.M. van der Heijden
Prof Dr H. Meurs
P. Jittrapirom, Dr tech. (email@example.com)
Note: This is an open access article distributed under the terms of the
Creative Commons Attribution License, which permits unrestricted use,
distribution, and reproduction in any medium, provided the original author and
source are credited.
Jittrapirom, P., Marchau, V., van der Heijden, R., Meurs, H. (August, 2018).
Dynamic Adaptive Policymaking for Implementing Mobility As a Service
(MaaS). Available at Radboud Repository
For more information on the SCRIPTS project, visit https://goo.gl/e6Nr2j
Mobility-as-a-Service (MaaS) is an innovative transport concept that combines a range of transport
modes and services to provide a user-orientated service via a single interface. Since its emergence,
MaaS has drawn increasing interest within and beyond the transport sector for its potential as an
innovative and potentially effective solution to urban transport problems. However, the implementation
of MaaS is surrounded by uncertainties concerning various aspects, such as on technological feasibility,
future demand and willingness of crucial stakeholders to cooperate. These uncertainties can prevent
large-scale implementation of the concept from taking place.
In this paper, an adaptive approach is proposed, which allows policymakers to create policies that are
more robust for uncertain future situations and which can be adapted as the future unfolds and
uncertainties are resolved. In particular, a Dynamic Adaptive Policymaking (DAP) is currently being
developed for implementing MaaS for the Dutch city of Nijmegen. The study is based on a desktop
research and has been produced by discussion among a small group of experts. Its outcomes are
presented in an initial plan, which a real-world project could be based upon, and an alternative planning
approach could be designed to handle uncertainty.
This research is part of SCRIPTS: Smart Cities Responsive Intelligent Public Transport Systems, a
SURF (Smart Urban Regions of the Future) project, carried out by the TU Eindhoven, TU Delft, Radboud
University and HAN University of applied sciences. It is sponsored by NWO (The Netherlands
Organization for scientific research in close collaboration with a large number of social partners. The
Many cities around the world are facing challenges in managing their transport systems. Urban
citizens do not only profit from mobility opportunities, but they also suffer from the negative
externalities of road transport, such as poor air quality, extended travel time, traffic accidents
and congested road spaces (Edwards & Smith, 2008; Hayashi, Doi, Yagishita, & Kuwata, 2004;
Taipale, Kaarin. Fellini, Claire. Le Blanc, 2012; Zavitsas, Kaparias, Bell, & Tomassini, 2010).
The existing trend in digitalisation, however, provides new opportunities for improving the
performance of the transport sector and limiting the negative external effects (CIVITAS, 2016;
Holmberg, Collado, Sarasini, & Williander, 2015), mainly by enabling innovative transport
services. One such opportunity is provided by the concept of Mobility-as-a-Service (MaaS).
MaaS is a transport distribution concept which is focused on passenger transport, and which
combines a range of transport modes and services to provide a user-orientated transport
solution via a single interface. The service can be provided in exchange for a pay-as-you-go
fee or a monthly subscription, similar to a mobile phone (Hietanen, 2014). MaaS is therefore
expected to provide an attractive and smart solution to solve the problem of urban mobility
(Nemtanu, Schlingensiepen, Buretea, & Iordache, 2016). It could cause a potential paradigm
shift in the transport system that can ‘free’ the users from any potential mode-specified sunk
costs, such as car ownership or annual public transport subscription fees, which potentially
‘lock’ users into using certain modes. Instead, through MaaS, users can flexibly combine the
available modes to best fit their changing travel needs via a digital platform or a virtual
marketplace that mediates mobility supply and demand (Meurs & Timmermans, 2016). MaaS
can be considered as the next step in integrated transport or multimodal mobility, through its
emphasis on the use of digitalisation and inclusion of the ‘business dimension’. It also provides
an opportunity to connect transport services with services from other sectors, such as tourism
and entertainment (Finger, Bert, & Kupfer, 2015).
MaaS’ potential has drawn increasing interest within and beyond the transport sector. Its
proponents believe that the implementation of the concept on a large scale could have
significant implications for the transport system in several ways. For instance, MaaS could
change the way public authorities provide and subsidise their transport services, as it offers an
option to improve customer experience and capabilities in conventional public transport
services (Hensher, 2017). This makes it possible to provide accessibility-based concessions
instead of mode-based concessions which, in turn, has the potential to affect the efficiency of
public spending on transport. MaaS is believed to have the potential to increase the use of
public or shared transport facilities by offering a high-level of convenience which would
persuade drivers to give up or to reduce their use of private vehicles (Holmberg et al., 2015).
As such, the implementation of MaaS could contribute to the performance of the transport
system by reducing congestion, decreasing the need for parking space, and enhancing the
level of accessibility (European Comission, 2016). Moreover, the potential opportunities offered
by MaaS to business and public organisations to connect their services, such as in the cases
of health care and tourism, to the new transport services make it highly attractive to these
sectors (Finger et al., 2015).
However, these promises are based on highly uncertain assumptions. For instance, the
potential of MaaS to reduce road and parking congestion heavily depends, among others, on
the willingness of users to adopt and share services. Otherwise, such tailored mobility solutions
can even lead to “more vehicles on the road” (Mulley, 2017). Another uncertainty involves the
nature and feasibility of a sound business model for the MaaS-platform owners. Results from a
pilot project in Ubigo, a transport broker service in Gothenburg (Sweden), showed a reduction
of the participant's’ car use, a desirable societal goal, but this also lowered the revenue of the
platform owner, because Ubigo was unable to price the public transport trip higher than the
market rate. Thus it relied on making profits from the utilisation of other modes in the package,
such as taxi, bicycle sharing, car sharing, and car rent (Sochor, Strömberg, & Karlsson, 2015).
Another issue regarding the implementation of MaaS concerns the inherent risks due to the
increased use of centralised ICT-based transport services, which enable parties to have access
to essential data and to communicate with users. These ICT requirements affect security risks
and privacy issues (Finger et al., 2015). Next, the willingness and the preferences of transport
service providers and other service providers to cooperate with platform owners is uncertain
(Polis, 2017). In addition, there is uncertainty regarding equity in access to mobility through
MaaS, as this can easily be influenced by a platform aggregator (Jittrapirom, Caiati, et al.,
2017). Consequently, MaaS might be implemented only for those areas and user groups which
are beneficial from a transport operator point of view, not from a societal point of view. Finally,
there are also uncertainties that restrict MaaS implementation on a large scale, such as
scalability of the concept (Mulley, 2017) and the contributions of MaaS to transport system
goals as a whole (Zmud, Goodin, Tooley, Williams, & Affleck, 2018).
In order to reduce these (and other) uncertainties, alternative MaaS schemes have been
implemented in other areas of the world. Among these alternative versions, include pilot
projects that operated temporally within a defined period such as in Tampere (Finland). Others
are ongoing operational schemes, such as Kyyti MaaS (Helsinki), Whim (Amsterdam, Antwerp,
Helsinki, and West Midlands UK), and MyCicero (Italy). In addition, there are several projects
in the planning, including pilot projects in the Netherlands (Limburg, Utrecht-Leidsche Rijn,
Amsterdam Zuidas, Rotterdam-the Hague Airport, Eindhoven, Groningen-Drenthe, Twente,
and Nijmegen), Kyyti (Nashville, USA), Singapore, and Ghent (Belgium).
Since the number and size of these projects are still limited, and cross-border operation is still
lacking, the systematic evaluation of outcomes and the drawing of generic lessons learnt can
be challenging. Consequently, no real answers have been formulated yet on the question of
how to cope with the aforementioned uncertainties in the context of the large-scale
implementation of MaaS.
Traditional approaches for dealing with the type of uncertainties mentioned above are based
on the use of predictive models and/or scenario-based approaches. These ‘predict and act’
approaches are, however, insufficient for the context of MaaS implementation, as being based
on explorations of the past mechanisms which might easily change in the future. Moreover,
MaaS does not have such a history. Most of the uncertainties related to MaaS implementation
can therefore only be reduced through an approach based on monitoring and adaptation. The
approach proposed here to cope with the uncertainties related to implementing MaaS is the so-
called Dynamic Adaptive Policymaking (DAP) approach. DAP allows policymakers to create
policies that are more robust for uncertain future situations and can adapt as the future unfolds
and uncertainties resolve. This DAP approach is applied for a specific case: the implementation
of a MaaS system for the Dutch city of Nijmegen. The application of DAP in this study
demonstrates the possibilities to systematically cope with uncertainties in policy making
regarding the implementation of MaaS.
In addition to the aforementioned literature, a growing number of academic publications deal
with a variety of aspects, such as conceptual models underlying MaaS (Nemtanu et al., 2016),
a review and definition of the crucial concepts (Jittrapirom, Caiati, et al., 2017), the evaluation
of existing MaaS schemes (Kamargianni, Li, Matyas, Schafer, & Schäfer, 2016), the formulation
of challenges for governance (Pangbourne, Stead, Mladenović, & Milakis, 2018), and business
model development (Sarasini, Sochor, & Arby, 2017), a possible assessment framework
(Wong, Hensher, & Mulley, 2017), and likely development scenarios (Smith, Sochor, &
Karlsson, 2017), the exploration of possible applications to the tourism sector (Signorile,
Larosa, & Spiru, 2018) and potential travel behaviour impacts (Sochor, Karlsson, & Strömberg,
2016). This illustrated that more integrated studies on processes of implementation are scarce.
On the other hand, publications on the application of DAP have been reported, such as for
Innovative urban transport solutions (Marchau, Walker, & van Duin, 2008), climate change
(Rahman, Walker, & Marchau, 2008), road pricing (Marchau et al., 2010), Airport planning
(Kwakkel, Walker, & Marchau, 2010), infrastructure planning (Wall, Walker, Vincent, & Bertolini,
2015), and autonomous vehicle driving (Walker & Marchau, 2017).
By combining the focus on integrated MaaS implementation and DAP, this study makes a
unique contribution to the scientific analysis of planning for MaaS. Moreover, the findings and
the case study presented in this paper will be useful to practitioners coping with the
uncertainties surrounding this innovative service concept.
The structure of the remainder of this paper is as follows. In section 2, we present a framework
for policy analysis and provide a classification of uncertainty associated with each entity in the
framework. Next, we describe the DAP framework that supports decision makers in dealing
with high levels of uncertainty in section 3. We then apply the framework in our case study in
Sections 4 and 5 and conclude the paper in Section 6.
2 A framework for policy analysis and uncertainty
Policymaking involves identifying a set of measures or interventions to a system, with an aim
to gain desirable outcomes. In Figure 1 below, a framework has been proposed by Walker
(2000), for understanding policymaking. The core of this framework is the system domain (R),
which in our case, is the urban passenger transport system. We can define the boundaries and
structure within the system domain as follows: the main entities are the subjects of
transportation (people), the means of transportation (vehicles), and related infrastructures
(roads, rail within an urban environment). Their mutual interactions produce outcomes of
interest (O), which in this case can be levels of accessibility to urban spots, economic
performance, energy consumption, the level of emissions, the level of congestion, and the level
of traffic safety (e.g. casualties, injuries). Policymakers value these outcomes based on their
subjective preferences or weighing (W). Subsequently, they evaluate these outcomes against
their set goals or whether the perceived problems are resolved.
Policies (P) and external forces (X) are two types of influence that act on the system domain.
The policies are a set of actions policymakers control, such as the providing of legislation for
public transport operations or constructing additional bicycle lanes. In contrast, the external
forces are beyond the reach of policymakers. Examples of such external forces in the field of
urban transport are e.g. population demographics, climate change, technological
developments, and economic developments.
Figure 1. A Policymaking framework (Walker, 2000)
In this framework, different levels of uncertainty can be distinguished for different locations in
the framework above (Walker & Marchau, 2017). For example, regarding external forces (X),
the uncertainty in (inter)national economic developments is considered high while the ageing
of society might be regarded as rather certain. Another example involves the impacts of policies
on the outcomes of interest through the system (O). For some policies, the impacts can be
rather well predicted (e.g. changing parking fees) while for other parking policies (e.g. Park and
Ride schemes) this seems more difficult. Walker et al., (2010) introduce a typology of
uncertainty based on four levels of uncertainty (see Figure 2 depicting the gradual transition of
a level of uncertainty from complete certainty (left) to total ignorance (right)).
Level 1 uncertainty is often treated through a simple sensitivity analysis of transport model
parameters, where the impacts of small perturbations of model input parameters on the
outcomes of a model are assessed. Level 2 uncertainty involves uncertainty that can be
described adequately in statistical terms. In the case of uncertainty about the future, Level 2
uncertainty is often captured in the form of either a (single) forecast (usually trend based) with
a confidence interval or multiple forecasts with associated probabilities. However, for Levels 3
and 4, it becomes difficult to predict the future using a probabilistic approach as only the
boundaries of the future are assumed to be known (level 3) or even unknown (level 4).
Figure 2: The progressive transition of levels of uncertainty from complete certainty to total
ignorance (Adapted from (Walker, Marchau, & Kwakkel, 2013)).
As argued, in the case of MaaS, the level of uncertainty surrounding its implementation is high.
These uncertainties can now be interpreted in terms of the levels on Figure 2. Firstly,
uncertainty associated with the external forces (X):certain forces, such as population
development, can be forecasted using past data with some accuracy (Level 1), whereas other
forces, such as the national economic development, acceptance of shared services, and
attitude with respect to car ownership are harder to predict accurately (Level 2 or 3). Secondly,
the complexity arising from the domain of MaaS (R): since urban transport is known to be a
highly complexed system, mainly due to the interconnectivity between the entities within it
(Kölbl, Niegl, & Knoflacher, 2008; May, 2003), certain transport policy measures can bring
about unintended effects that worsen the overall performance of the system (ADB, 2009; IET,
2010; Jittrapirom, Knoflacher, & Mailer, 2017; Pojani & Stead, 2015). This uncertainty is
associated with Level 4 uncertainty. Thirdly, practitioners have to deal with limited knowledge
about the overall effects of MaaS to the urban transport system (O), and user and stakeholders’
acceptance. It may be possible to speculate about likely outcomes from analogies of other
sectors, such as hospitality in Airbnb or within the retail sector from eBay or Amazon. However,
the speculation of this future is likely to have a limited level of accuracy at best and the opinions
from stakeholders and the scientific community on this future are polarised. This uncertainty
also links to Level 4. Finally, the valuation of the outcome (W) by decision makers is a serious
issue. Differences in valuation may be forecasted with some certainty, but the assumptions on
preferences and the weighting of effects can be influenced by other factors that have a high
level of uncertainty, such as public mood at the time of valuation. Hence, here we deal with
uncertainty at least at Level 4.
3 Dynamic adaptive policymaking (DAP)
The key idea to cope with the variety of uncertainty, notably Level 3 and 4 uncertainties, is to
move away from developing a static plan that will work well for one or more specific futures,
and in its place, construct a dynamic plan that is flexible, adaptable and perform well across the
full range of plausible futures (including surprises). With this awareness, Walker, Rahman, &
Cave (2001) developed a Dynamic Adaptive Policymaking (DAP) scheme which is based on
the Assumption-Based Planning (Dewar, 2002; Dewar, Builder, Hix, & Levin, 1993). The DAP
has since then been further elaborated and applied (e.g. Kwakkel et al., 2010; Van der Pas,
Walker, Marchau, van Wee, & Kwakkel, 2013). This scheme enables policymakers to deal with
the uncertainties surrounding the policy formulation process right away, instead of waiting for
information to become available. DAP is based on the recognition that a full reduction of
uncertainty is unattainable. Instead, it focuses on utilising future knowledge in making a robust
policy that is prepared to cope with (uncertain) vulnerabilities and opportunities. It emphasises
the importance of creating a policy framework that allows policy to be adapted and changed in
accordance with information gained and feedback received.
The DAP consists of two phases; 1) a design phase and 2) an implementation phase. In the
first phase, the dynamic adaptive policy, monitoring program, and various pre- and post-
implementation actions are formulated. The latter phase consists of the operationalising of the
policy, the monitoring of its performance, and the implementation of (ex-ante developed)
adaptation actions if necessary. The first key notion of DAP is “Vulnerabilities”: events that can
reduce the impact of a policy to a point where the policy is no longer successful. The second
key notion is “Opportunities”: events that can enhance or accelerate policy success.
The planning phase of DAP consists of five steps; the first and second steps are identical to the
traditional policy formulation, while the rest of the steps are unique to DAP. Figure 3 depicts the
five steps with a summarised description below.
Step I: Stage-setting step – involves the traditional starting activities in policymaking,
such as specifying objectives, a definition of success, constraints, and available policy
Step II: Assembling a basic policy – consists of selecting a preferred, initial policy to
be implemented and identifying the required conditions for the basic policy to be a
Step III: Increasing the robustness of the basic policy – involves identifying
vulnerabilities and opportunities of the selected policy, together with their associated
likelihood and actions to be implemented with the basic policy at t = 0 to decrease
unfavourable or amplify favourable effects. These actions are associated with specified
types of event and likelihood. For example, the Mitigating Action (M) is to reduce a
Certain Vulnerability, with the exception of the Shaping Action, an action taken to
control the likelihood of a vulnerability or an opportunity.
Step IV: Setting up the monitoring system - includes defining signposts to track
information and associated triggers, or the critical values of signpost variables, which,
once exceeded, will activate actions to change the policy to ensure the system is kept
moving as planned.
Step V: Preparing the trigger response – comprises the specification of a set of
responsive actions to be taken when a trigger level is reached after the basic policy is
implemented (at t > 0): Defensive Action (DA) to preserve the policy; Corrective Action
(CA) to adjust the basic policy; Capitalizing Action (CP) to take advantage of
opportunities arise, and Reassessment (RE) to re-evaluate or revise the whole basic
After the formulation of the dynamic adaptive policy is completed, the DAP proceeds from the
designing phase to the implementation phase; the basic policy is implemented together with
prior actions specified in Step III and the monitoring system specified in Step IV. The adaptive
process is suspended until a trigger value is reached and a responsive action is activated. In
certain cases, the responsive actions may not be sufficient to support the basic policy and the
basic policy needs to be revised altogether. In such cases, the experience and information
gained from setting up the initial adaptive policy can provide valuable input to the subsequent
In the next section, this DAP scheme is applied to develop an adaptive policy for implementing
a MaaS-concept in the city of Nijmegen, which is located in the Netherlands. This application
is a simplified example to illustrate the potential of DAP in this context.
Figure 3: DAP process (Walker et al., 2013)
4 Case study: Nijmegen city, the Netherlands
Nijmegen is a city in the province of Gelderland, situated in the eastern region of the
Netherlands. As of April 2017, the city had about 170,000 inhabitants and it is surrounded by
various suburban municipalities. Nijmegen is also adjacent to the city of Arnhem (15
kilometres), which has about 150,000 inhabitants, as of 2017. The proximity between the two
cities makes them often seen as twin cities, including a populated sub-urban in-between area
III. Increasing the Robustness of the Initial Plan
Shaping actions (SH)
Mitigating actions (M)
Hedging actions (H)
V. Preparing the trigger responses
Corrective actions (CA)
Defensive actions (DA)
Seizing actions (SZ)
IV. Setting up the monitoring system
Capitalizing actions (CP)
Exploiting actions (EZ)
Definition of success
I. Stage Setting
Necessary conditions for success
II. Assembling an Initial Plan
Nijmegen attracts a high amount of traffic, which is generated by its population and employment
in the area. The major employers are higher education institutions, three major hospitals, and
various service and production industries. The daily incoming traffic consists of commuters,
students, business people, and visitors to hospitals and city shopping centres. The outgoing
traffic is also mainly employment and education commuters to the surrounding city and the
western Dutch urban area in the Randstad region (City of Nijmegen, 2016).
Figure 4: Public transport network of Nijmegen (City of Nijmegen, 2016)
Public transportation plays an essential part in the city’s transport system (Figure 4). The main
railway (Nijmegen Central Station) provides intercity and regional connections between
Nijmegen with destinations such as various cities in the Northern Provinces, The Randstad
cities including Amsterdam Schiphol Airport in the West, various cities in Southern Provinces
and Maastricht, Antwerp and Brussels in the South and the German border town Venlo. It
facilitates approximately 45,000 passengers on average daily. The city has four other local
stations, one of which serves the Heijendaal district (University train station) that is largely
occupied by Radboud University, Radboud UMC (academic hospital) and HAN University of
Applied Sciences. Similarly to other cities in the Netherlands, Nijmegen has a high proportion
of cycling traffic, with over 24% of daily traffic. Nevertheless, private cars have the highest trip
proportion, with 41% (driver and passenger combined), whereas public transport accounts for
9% of all daily trips. This figure is slightly different if we consider only trips below 5 kilometres.
For that distance category, cycling trips have become dominant (37%), followed by car trips
(30%), and walking (28%). Public transport trips represent a meagre proportion of only 3-4%,
although 89% of the city’s population lives within 300 metres from a public transport line
(Figures from 2014 from City of Nijmegen, 2016).
Nijmegen city is currently experiencing increased transport congestion during rush hours. Long
queues of car traffic can be regularly observed around the two city bridges that link the city to
its surrounding area to the north. Additionally, there are long queues of bus and train
passengers for services between the central station and Heijendaal district during the peak-
hour periods as well. Breng, the local bus operator, has commissioned an exclusive bus rapid
transit service between the station and Heijendaal district to resolve the issue, but the problem
persists. These congestion problems are expected to intensify, as the residential area north of
the city (De Waalsprong) is being further developed, and the number of students enrolled at the
higher education institutions at Heijendaal continues to increase (in 2018, about 45.000
students take a study in the direct surroundings of the Heijendaal train station). Also, there is a
decrease in the percentage of bus patronage in certain lines, which has put a strain on the bus
operator to maintain the profitability of the operation of its 35 bus lines.
4.2 DAP for implementing MaaS in Nijmegen city
In this section, a possible adaptive policy for implementing MaaS in Nijmegen is developed from
the perspective of the local authority (i.e. the municipality of Nijmegen). The main aim here is
to demonstrate how DAP can be used to formulate an implementation plan for a MaaS scheme.
This dynamic adaptive plan is neither exhaustive nor finalised. It is simplified and based on a
desktop study and discussions among a group of experts. Only a selection of items is included
here and should be seen as a simplified first draft of an adaptive plan to initiate discussion with
stakeholders, who will enhance the sophistication and relevance of the plan through a
participatory process. We derived the planning items from related transport policy documents
(e.g. City of Nijmegen, 2017; CIVITAS, 2016; European Commission, 2014; Polis, 2017) and
literature on MaaS (e.g. Finger et al., 2015; Lund, 2017; Mulley, 2017; Sochor et al., 2016).
Step I Setting the Stage
We started the planning process by identifying the major transport policy objectives for
Nijmegen city from the official documents of the municipality, which outline the city’s goals on
mobility. We also reviewed documents on the related policy areas, such as climate and energy.
Overall, the local authority aims to maintain the levels of accessibility, reliability, and safety of
its transport system from the perspectives of visitors, residents and economic vitality of the
inner city. To reach these aims, the city authorities focus on a larger share of clean and
sustainable transport, such as bicycle and public transport, in the urban mobility patterns. In
particular, Nijmegen aims to increase the levels of its accessibility and road transport safety
(City of Nijmegen, 2017). Moreover, it aims to work with the neighbouring municipalities and
the industry to stimulate the production and sale of clean fuels and on working toward
sustainable urban transport (City of Nijmegen, 2018). The derived objectives are shown in
Table 1: City of Nijmegen’s Urban transport policy objectives
Ensure accessibility to key destinations and services for all citizens
Promote sustainable mobility, better use of resources and reducing negative impacts
of transport sector (noise, emissions, energy consumption) on the environment
Improve safety and security of road transport
Reduce road and parking congestion
Note: Extracted from (City of Nijmegen, 2016, 2017, 2018)
The definition of success is related to the defined objectives and involves the specification of
desirable levels of outcomes to achieve (e.g. desired levels of accessibility to key destinations
and services for all citizens and levels of public transport use, car-ownership, congestion, and
emissions). There is a number of available policy options for realising these objectives; such
as increasing road capacity, improving the existing public transport services, promote
sustainable transport, such as cycling and walking and/or introducing innovative transport
services such as MaaS. Given these available options, there are several related constraints,
for instance, the spatial constraint of the city resulting from its location next to the Waal river,
the budgetary limit of the municipality that may prevent a choice for any costly large-scale
project, and the public and political acceptance from the locals.
Step II Assembling a basic MaaS-policy
In this step, we identified the preferred policy option and the associated conditions for success.
Given a range of options available, MaaS can be a promising policy as it can improve the level
of accessibility to key destinations by providing a customised door-to-door mobility service for
travellers toward these destinations. It can also improve the match between demand and supply
and as such contribute to an improvement in the sustainability of the transport system by
enhancing its efficiency. Additionally, the hyper-convenient transport service in MaaS is likely
to increase public transport patronage and attracting car drivers away from their personal
motorised vehicle. Hence, MaaS would reduce the negative impacts of the transport system on
the environment and decrease the demand for road and parking spaces.
There is a range of policy measures that can be adopted to implement MaaS (See Table 2
below for available policy measures to implement MaaS as proposed and studied in the
literature). For our case study, we assume that the policymakers decided to implement living
labs or a showcase of pilot projects, as the most promising policy (i.e. least uncertainties).
This measure will allow actors involved to build up experience in operating this innovative
transport service and gaining confidence from potential users and stakeholders. We
hypothetically assumed the period of the pilot project to be 2-year for the subsequent steps of
the planning process.
Table 2: A selection of available policy measures to implement MaaS
Experiments and implementation
Implementing living labs or a showcase of pilot projects to create experiencing
and learning through cooperative pilot activities
Ensure related infrastructures necessary to operate MaaS (transport and digital)
are in place or focus MaaS implementation on the geographical area with good
Planning and policy
Reform transport planning and policy practice to promote integrated/service-
oriented mobility planning with a focus on sustainability, equity and consumer
Link MaaS operation as an objective with other governmental policy (such as
innovation, health, business)
Governance and regulation
Establish a qualified leadership (a statutory body) to oversee MaaS development
and identify a clearer division of the new roles and responsibilities of the various
Ensure that rules, regulation, policies, actions and mindset of both individuals
and organisations contribute to innovation supporting the development and
implementation of MaaS (e.g. open data policy)
Next, we compiled a set of necessary conditions for the success of the selected basic policy by
reviewing related literature. A selection of these conditions is listed in Table 3.
Table 3: Necessary conditions for success of the selected basic policy (i.e. implementing a 2-
year MaaS pilot project)
Condition of success
a) Collaboration, cooperation, and coordination based on trust
and shared visions between key actors, stakeholders, and a
(EC, 2015; Finger et al., 2015;
Holmberg et al., 2015; Karlsson et
al., 2016; Sochor et al., 2017)
b) Availability and standardisation of data on mobility and other
(Finger et al., 2015; Lund et al.,
2017; Sochor et al., 2017)
c) Strong and continuous financial support from related
(Lund et al., 2017)
d) Attractive business opportunities for potential actors
(Lund et al., 2017; Smith et al., 2017)
e) Provision of appropriate physical infrastructure
(Karlsson et al., 2016; Lund et al.,
f) Suitable regulation regarding data security and privacy
(Finger et al., 2015; Lund et al.,
g) Availability of enabling technology
(CIVITAS, 2016; Lund et al., 2017;
h) Apparent added value of MaaS from customer perspective
e.g. personalised service, easy access, easy payment, etc.
(Lund et al., 2017; Sochor et al.,
Step III Increasing the robustness of the basic policy
The robustness of the basic policy is considered in this step by examining each condition for
success for its inherent vulnerability or opportunity. Next, the certainty of the identified
vulnerability or opportunity is assessed. A number of actions are then put forward to improve
the policy’s robustness, protecting it from failing and/or enabling to take opportunities. A
summary is presented in Table 4. The columns 1-4 of this table present an overview of
Vulnerabilities (V) and Opportunities (O), related to the conditions of success identified in the
previous step, and whether these vulnerabilities and opportunities are certain or uncertain. The
conditions of success present are correlated to Table 3. Column 5 of Table 4 contains the
actions to be implemented together with the implementation of the basic policy (t = 0) to
increase the basic policy’s robustness.
Step IV and V defining a monitoring system and specifying responsive actions
In Step IV, a monitoring system is set up to keep track of the plan by observing any changes in
the identified conditions for success. The signposts and their trigger levels, as part of the
monitoring system, are shown in column 6 of Table 4, and the associated adaptive actions are
specified in the final column.
To elucidate the adaptation plan, we will highlight some elements of the table to demonstrate
the essence of the DAP scheme and its application to the case study. In the first row (a.
Collaboration between key actors, stakeholders, and platform operator), the lack of willingness
to act accordingly may be certain, as MaaS require participating actors, such as public transport
providers and the local authority to share data with and provide their services through a platform
operator. This can be equated to giving up their strong or a monopolistic position. A potential
lack of collaboration can derail the project. A possible Shaping Action (SH) to decrease this
threat is to secure support from critical actors through lobbying and to provide incentives to
actors to collaborate (e.g. some rewards or subsidy). Additionally, a Mitigating Action (M) is to
secure funding that can be used to subsidise the service, thus reducing the reluctance of public
transport providers to join the project. The associated signposts will monitor the level of the
collaboration between stakeholders (e.g. highly engaged, not interested, or hostile). Should
these levels drop below expectation, the need for adaptive actions will be triggered, such as
intensifying stakeholder engagement (CA), and/or reducing the scope of the offered Maas
service to minimise the impacts of absent stakeholders (DA).
Table 4: Vulnerabilities and opportunities in the basic policy to implement a MaaS pilot project with the 2-year period for Nijmegen city
Vulnerability (V) / Opportunity
Actions are taken at Time = 0
(Increase Basic Policy
Signpost Monitoring (begins at
Time = 0) and Trigger Events
Actions taken at Time > 0
between key actors,
Lack of willingness to
collaborate due to different
reasons, such as lack of trust,
or potential loss of operational
(SH) Lobby for support from critical
actors and actively involve relevant
stakeholders; Provide incentives to
actors to collaborate
(M) Secure funding for incentives or
subsidy to ensure service operation
Monitor: Level of collaboration and
actor feedbacks on plan
Trigger: Level of participation drops
below a predetermined threshold level
(DA) Adjust basic policy in
response to feedback
(CA) Intensify engagement with
stakeholders to increase
b. Availability and
related data, such as
Unavailability of necessary
data or inappropriate data
(SH) Consult actors on preferred
data format and establish a common
(H) Appoint an expert on data
management to support the pilot
Monitor: Feedback from actors; the
number of incompatible data
Trigger: Negative feedback from
platform integrator on the matter;
Number of incompatible data increase
above a set threshold.
(CA) Adjust data format per
feedback; Establish data
c. Strong and
support from related
Disruption of financial support
from public authority during
the pilot project
(SH) Secure funding for the pilot for
the planned period; manage actors’
expectations on the pilot
(H) Design the pricing of the pilot to
minimise dependency on external
funding; diversify the source of
Monitor: support from related public
Trigger: level of support fall below set
(CA) Adjust authority’s expectation
on the pilot; lobby for a permanent
(DA) adjust the level of service to a
minimum to reduce related cost
(RE) discontinue the service
Additional financial support
during the pilot project
(SH) Report outcomes of the pilot
project to relevant organisations
Monitor: Interests from the related
Trigger: Interests generates additional
(CP) Expand area of coverage or
increase level of services
d. Attractive business
MaaS offers a highly attractive
MaaS offer the below-than-
(SH) Involve business sector in
designing MaaS; Designing the pilot
to minimise risks to actors and
(SZ) Report success to attract
(H) Secure additional funding to
subsidise actors and stakeholders
Monitor: Feedback from actors and
stakeholders; actors’ revenue and
Trigger: Level of satisfaction of actors
and stakeholders drop below a
predetermined threshold level;
operational costs exceed the revenue
above-set threshold level
(CA) Adjust basic policy in
response to feedback
(RE) Redesign MaaS service’s
(DA) Scale down operation to
Note: (SH) = Shaping Action; (H) = Hedging Action; (M) = Mitigating Action; (SZ) = Seizing Action; (EZ) = Exploiting Action; (CA) = Corrective Action; (DA) = Defensive Action; (CP) = Capitalizing
Action; (RE) = Reassessment
Table 4: Vulnerabilities and opportunities in the basic policy to implement a MaaS pilot project with the 2-year period for Nijmegen city (continued)
Vulnerability (V) / Opportunity
Actions are taken at Time = 0
(Increase Basic Policy
Signpost Monitoring (begins at
Time = 0) and Trigger Events
Actions taken at Time > 0
e. Provision of
Lack of appropriate physical
infrastructure prevents MaaS
from fully functioning (e.g. lack
of shared bicycle facility)
(SH) Identify gap between desirable
and existing infrastructures, discuss
with relevant parties to address the
(M) Adjust MaaS design to optimise
Monitor: Gap between desirable and
available infrastructures; the level of
dependency on expected infrastructure
Trigger: a wide gap between desirable
and available infrastructure; high level
of dependency on expected
(CA) Adjust MaaS design if
possible. If not, reassess the
f. Suitable regulation
regarding data security
Lack of suitable regulation on
data security and privacy
(SH) Adopt existing practices (e.g.
General Data Protection
Regulation) and adjust to fit MaaS;
appoint a data security expert to
support the pilot
(H) Enact internal agreements on
the subject for the pilot between
actors and other stakeholders
Monitor: the presence of suitable
regulation and their relevance to MaaS
Trigger: Lack of suitable regulation
(CA) Establish temporary protocol
on data security that is the binding
responsibility of all parties.
g. Reliable technology
to support operation
Unavailability of key
(SH) Include actors with required
capability in the pilot project and
secure their support
(H) Shortlist alternative technology;
Prepare additional funding to
Monitor: secure technical support from
Trigger: Unavailability of key
(CA) seek support from other pilots
or use alternative technology
(RE) Postpone the pilot project
(SH) Deploy proven technology with
high-level technical support. The
soft launch of the system to iron out
any possible bugs,
(H) Prepare non-digital contingency
plans, such as a telephone call
centre with a manual vehicle
Monitor: Operations report, Customer
feedback; system operation status
Trigger: Level of negative feedback
and failure incidents rise above
(CA) Conduct investigation and
(DA) Roll back to operate as a non-
digitalisation integrated transport
h. Apparent added
value of MaaS for
Customers perceive a high
added value of MaaS
Customers do not perceive
any added value of MaaS
(SH) Design MaaS service focusing
on the customers; Pre-launch MaaS
to test with a small group of users
(EZ) Expand MaaS pilot’s coverage
(H) Launch marketing campaign to
highlight the values of MaaS
Monitor: customer feedbacks; service
adoption and usage
Trigger: Levels of satisfaction or
service adoption and usage fall below
(CA) Adjust service design
(DA) Readjust expectation of
(RE) Reassess system design
Note: (SH) = Shaping Action; (H) = Hedging Action; (M) = Mitigating Action; (SZ) = Seizing Action; (EZ) = Exploiting Action; (CA) = Corrective Action; (DA) = Defensive Action;
(CP) = Capitalizing Action; (RE) = Reassessment
5 Implications for managerial practice
The findings of this study have two apparent implications for the practice.
First, the wide range of identified planning elements related to MaaS suggests that the concept needs
to be encompassing. The aim to realize an encompassing concept from the start is optimistic. This
optimism stems from the fact that the concept is still in its early stage of development and that only a
limited number of wide-scale examples of application in practice are known. The (mainly theoretical)
advantages influence the present hype on what MaaS can achieve and may seem to contribute
significantly to the high expectations of the concept. However, as argued in this paper, to implement
MaaS in a comprehensive way poses a serious challenge. The comprehensive nature of MaaS requires
that an extended and complex configuration of services is offered to meet all the travellers’ expectations,
which leads to extra but also an essential condition for success. This, in turn, might significantly increase
the amount of resources required for implementation. On the other hand, it may be expected that the
complexity of MaaS decreases or will be more manageable as the concept further develops in practice
with more real-life examples and experiences from the practise. Alternatively, a reduction of scope in
the start-up phase of MaaS might be achieved by planning for a MaaS service with a limited focus on
solving a specific challenge, such as to provide an improved access to a newly developed residential
area with low provision of parking space.
Second, the application of DAP to MaaS planning shows that the framework offers a structured way of
thinking about the various types of uncertainty related to the innovative MaaS concept. DAP can be
used to systematically plan the implementation of the concept in a specific geographical area and to
identify actions that can increase the robustness of the plan, hence increases the plan’s likelihood to
succeed. In addition, the DAP can aid resource allocation through more in-depth assessment of the
level of uncertainty of each required condition for success. Irrespective of its promising prospects, the
DAP still requires further development, as mentioned in the following section.
6 Conclusion / Discussion
In this paper, we explained and applied the Dynamic Adaptive Policymaking (DAP) approach in the
context of an initiative taken by local stakeholders for implementing an initial MaaS-system for the Dutch
city of Nijmegen. According to the DAP-scheme, we detailed the objectives to implement MaaS and the
definition of success. Also, we identified a range of possible policies and likely constraints. We then
explored the choice for the implementation of a living lab or a showcase of pilot projects as our
presumptive example of a basic policy. Various necessary conditions for success for this basic policy
were identified, including collaboration between key stakeholders and platform operators, the quality of
the applied ICT services and the availability of standardised data and the specification of a realistic
business plan. The opportunities and vulnerabilities of each condition for success and their levels of
uncertainty are made explicit. We then proposed a number of actions that should be taken right away
from the beginning to increase the robustness of the basic MaaS policy, such as:
Secure the support from critical actors, actively collaborate with stakeholders and provide
incentives for them to collaborate (Shaping Action),
Secure the funding for incentives or subsidies to ensure service operation for the starting period
Focus on a soft launching of the system enabling the handling of any possible bug preceding
an official and more large-scale launch (Hedging Action).
Finally, the contours of a monitoring system were specified to trigger future adaptations of the basic
policy in reaction to increase in knowledge, e.g. related to the acceptance and performance of the MaaS
pilot. For instance, the stakeholders’ levels of acceptance or the level of ridership, and the cost-revenue
ratio are monitored, and responsive actions are prepared to be implemented in case trigger-events
occur. For example, a high level of acceptance by travellers may trigger an extension of a MaaS scheme
as a Capitalizing Action.
We demonstrated in this paper how DAP can offer an alternative transport planning method, which adds
to traditional approaches an explicit process-oriented focus on uncertainty and the inclusion of the
concept of adaptation. The DAP method assumes that uncertainty is non-binary and inevitable. It
enhances the robustness of a given policy plan, recognizing the surrounding dynamic changes, thus
increasing the likelihood of its success. However, so far DAP has been applied mostly by researchers
in hypothetical case studies.
Consequently, there remain important challenges in applying the adaptive planning approach in practice
(Bosomworth et al., 2017). These challenges include difficulties in using DAP to deal with complex and
contested issues, establish trigger points for a complex system, and to take into account the institutional
and governance implications of applying DAP. Elaborating on these challenges can improve the DAP
approach and is needed to ensure its successful application in the real world.
Up until now, the simplified DAP scheme presented in this paper has been the result of a desktop study
and discussions among a limited group of experts, although the policy triggers from the city of Nijmegen
is realistic. The preliminary analysis in this study has apparent implications for future research regarding
this policy initiative. As such, the findings in the study act as a starter for developing a real-world MaaS-
DAP in the Nijmegen region, or for other case studies. The subsequent studies should at least
incorporate opinions of experts from a wider field and organize a participatory process with involved
actors and stakeholders. Additionally, consideration should be given on how to cope and account for
different preferences and opinions from the involved stakeholders, who might have conflicts of interest.
The development of such a participatory MaaS-DAP will be challenging, but nevertheless it would enable
the plan to be more comprehensive and more implementation-ready.
ADB. (2009). Changing Course - A NEW Paradigm for Sustainable Urban Transport. Image (Rochester,
N.Y.) (Vol. 2010). 2008.
Bosomworth, K., Leith, P., Harwood, A., & Wallis, P. J. (2017). What’s the problem in adaptation
pathways planning? The potential of a diagnostic problem-structuring approach. Environmental
Science and Policy, 76(January), 23–28. https://doi.org/10.1016/j.envsci.2017.06.007
City of Nijmegen. (2016). Application for European Green Capital. City of Nijmegen.
City of Nijmegen. (2017). Mobility objectives of Nijmegen city. Retrieved August 26, 2017, from
City of Nijmegen. (2018). Budget 2018 Municipality of Nijmegen. Retrieved June 26, 2018, from
CIVITAS. (2016). Mobility-as-a-Service: A new transport model. Civitas Insight, 18, August, 10.
Dewar, J. A. (2002). Assumption-Based Planning: A Tool for Reducing Avoidable Surprise.
Dewar, J. A., Builder, C. H., Hix, W. M., & Levin, M. H. (1993). Assumption-based planning: a planning
tool for very uncertain times. https://doi.org/10.1017/CBO9780511606472
EC. (2015). Commission Decision: setting- up an expert group on Digital Transport and Logistics ('the
Digital Transport and Logistics Forum’). Brussels.
Edwards, T., & Smith, S. (2008). Transport Problems Facing Large Cities. NSW Parliamentary Library
Briefing Paper, (6).
European Comission. (2016). Horizon 2020 - Smart, Green and Integrated transport. Imporant Notice
on the Scond Horizon 2020 Work Programme, 2017(July 2016), 129.
European Commission. (2014). Towards a roadmap for delivering EU-wide multimodal travel
information, planning and ticketing services. Commission Staff Working Document, 1(9), 33–36.
Finger, M., Bert, N., & Kupfer, D. (2015). Mobility-as-a-Service: from the Helsinki experiment to a
European model? FSR Transport, 01(2015). https://doi.org/10.2870/07981
Hayashi, Y., Doi, K., Yagishita, M., & Kuwata, M. (2004). Urban Transport Sustainability : Asian Trends ,
Problems and Policy Practices. EJTIR, 4(1), 27–45.
Hensher, D. A. (2017). Future bus transport contracts under mobility as a service regime in the digital
age : are they likely to change ? Transportation Research Part A, 98, 86–96.
Hietanen, S. (2014). ‘ Mobility as a Service ’ – the new transport model ? Eurotransport, 12(2), 2–4.
Holmberg, P.-E., Collado, M., Sarasini, S., & Williander, M. (2015). Mobility as a Service: Describing the
IET. (2010). Rebound - Unintended consequences from transport policies and technology innovations.
Jittrapirom, P., Caiati, V., Feneri, A. M., Alonso-González, M. J., Ebrahimigharehbaghi, S., & Narayan,
J. (2017). Mobility as a Service : a critical review of definitions, assessments of schemes, and key
challenges. Urban Planning, 2(2).
Jittrapirom, P., Knoflacher, H., & Mailer, M. (2017). The conundrum of the motorcycle in the mix of
sustainable urban transport. Transportation Research Procedia, 25, 4869–4890.
Kamargianni, M., Li, W., Matyas, M., Schafer, A., & Schäfer, A. (2016). A Critical Review of New Mobility
Services for Urban Transport. In Transportation Research Procedia (Vol. 14, pp. 3294–3303).
Karlsson, I. C. M., Sochor, J., & Strömberg, H. (2016). Developing the ‘Service’ in Mobility as a Service:
Experiences from a Field Trial of an Innovative Travel Brokerage. Transportation Research
Procedia, 14, 3265–3273. https://doi.org/10.1016/J.TRPRO.2016.05.273
Kölbl, R., Niegl, M., & Knoflacher, H. (2008). A strategic planning methodology. Transport Policy, 15(5),
Kwakkel, J. H., Walker, W. E., & Marchau, V. A. W. J. (2010). Adaptive Airport Strategic Planning.
European Journal of Transport and Infrastructure Research, 10(3), 249–273.
Lund, E., Kerttu, J., & Koglin, T. (2017). Drivers and Barriers for Integrated Mobility Services.
Marchau, V. A. W. J., Walker, W. E., & van Wee, G. P. (2010). Dynamic adaptive transport policies for
handling deep uncertainty. Technological Forecasting and Social Change, 77(6), 940–950.
Marchau, V., Walker, W., & van Duin, R. (2008). An adaptive approach to implementing innovative urban
transport solutions. Transport Policy, 15(6), 405–412.
May, A. D. (2003). Prospects: A Decision Makers’ Guidebook. Brussels: European Commission.
Meurs, H., & Timmermans, H. Mobility as a Service as a Multi-Sided Market : Challenges for Modeling,
TRB 2016 annual meeting § (2016).
Mulley, C. (2017). Mobility as a Services ( MaaS ) – does it have critical mass ?, 1647(May).
Nemtanu, F., Schlingensiepen, J., Buretea, D., & Iordache, V. (2016). Mobility as a service in smart
cities. In Responsible entrepreneurship - vision, development and ethics: Proceedings of the 9th
International conference for entrepreneurship, innovation and regional development. June 23-24,
2016 Bucharest, Romania (pp. 425–435).
Pangbourne, K., Stead, D., Mladenović, M., & Milakis, D. (2018). The Case of Mobility as a Service: A
Critical Reflection on Challenges for Urban Transport and Mobility Governance. In Governance of
the Smart Mobility Transition (pp. 33–48). Emerald Publishing Limited. https://doi.org/10.1108/978-
Pojani, D., & Stead, D. (2015). Sustainable Urban Transport in the Developing World: Beyond
Megacities. Sustainability, 7(12), 7784–7805. https://doi.org/10.3390/su7067784
Polis. (2017). Mobility as a Service: Implications for urban and regional transport.
Rahman, A., Walker, W. E., & Marchau, V. (2008). Coping with Uncertainties About Climate Change in
Infrastructure Planning – An Adaptive Policymaking Approach, (November), 1–57.
Sarasini, S., Sochor, J., & Arby, H. (2017). What characterises a sustainable MaaS business model? In
ICoMaaS 2017 Proceedings (pp. 121–135).
Signorile, P., Larosa, V., & Spiru, A. (2018). Mobility as a service: a new model for sustainable mobility
in tourism. Worldwide Hospitality and Tourism Themes, 10(2), 185–200.
Smith, G., Sochor, J., & Karlsson, I. M. (2017). Mobility as a Service: Implications for future mainstream
public transport. Thredbo15 - International Conference Series on Competition and Ownership in
Land Passenger Transport.
Sochor, J., Arby, H., & Karlsson, M. (2017). The topology of Mobility as a Service: A tool for
understanding effects on business and society, user behavior, and technical requirements. 24th
World Congress on Intelligent Transportation Systems.
Sochor, J., Eckhardt, J., König, D., & Karlsson, I. C. M. (2016). Future needs and visions for Mobility as
a Service : Insights from European workshops, (October), 10–14.
Sochor, J., Karlsson, I. C. M., & Strömberg, H. (2016). Trying Out Mobility as a Service: Experiences
from a Field Trial and Implications for Understanding Demand. In 95th Annual Meeting of the
Transportation Research Board, Washington, D.C., January 10-14, 2016. Washington, DC.
Sochor, J., Strömberg, H., & Karlsson, M. (2015). Challenges in integrating user, commercial and
societal perspectives in an innovative mobility service. In 94th Annual Meeting of the
Transportation Research Board. Washington, D.C.,.
Taipale, Kaarin. Fellini, Claire. Le Blanc, D. (2012). Challenges and way forward in the urban sector.
Sustainable Development in the 21st Century (SD21).
Van der Pas, J. W. G. M., Walker, W. E., Marchau, V. A. W. J., van Wee, B., & Kwakkel, J. H. (2013).
Operationalizing adaptive policymaking. Futures, 52, 12–26.
Walker, W. E. (2000). Policy analysis: a systematic approach to supporting policymaking in the public
sector. Journal of Multi-Criteria Decision Analysis, 9(1–3), 11–27. https://doi.org/10.1002/1099-
Walker, W. E., & Marchau, V. A. W. J. (2017). Dynamic adaptive policymaking for the sustainable city:
The case of automated taxis. International Journal of Transportation Science and Technology.
Walker, W. E., Marchau, V. A. W. J., & Kwakkel, J. H. (2013). Uncertainty in the Framework of Policy
Analysis. In W. A. H. Thissen & W. E. Walker (Eds.), Public Policy Analysis: New Developments
(pp. 215–261). Boston, MA: Springer US. https://doi.org/10.1007/978-1-4614-4602-6_9
Walker, W. E., Marchau, V. A. W. J., & Swanson, D. (2010). Addressing deep uncertainty using adaptive
policies: Introduction to section 2. Technological Forecasting and Social Change, 77(6), 917–923.
Walker, W. E., Rahman, S. A., & Cave, J. (2001). Adaptive policies, policy analysis, and policy-making.
European Journal of Operational Research, 128(2), 282–289. https://doi.org/10.1016/S0377-
Wall, T. A., Walker, W. E., Vincent, A. W. J. M., & Bertolini, L. (2015). Dynamic Adaptive Approach to
Transportation-Infrastructure Planning for Climate Change : San-Francisco-Bay-Area Case Study.
Journal of Infrastructure Systems, 21(4), 05015004. https://doi.org/10.1061/(ASCE)IS.1943-
Wong, Y. Z., Hensher, D. A., & Mulley, C. (2017). Emerging transport technologies and the modal
efficiency framework: A case for mobility as a service (MaaS), 1–24.
Zavitsas, K., Kaparias, I., Bell, M., & Tomassini, M. (2010). Transport problems in cities. ISIS, (1).
Zmud, J., Goodin, G., Tooley, M., Williams, T., & Affleck, P. (2018). White Paper on Near-Term
Research Needs: Automation, Connectivity, Big Data, Mobility Service Models.