Mobility-oriented Agenda Planning as a Value-adding Feature for

Mobility as a Service

Felix Schwinger

1,2

and Karl-Heinz Krempels

1,2

Fraunhofer Institute for Applied Information Technology FIT, Aachen, Germany

Information Systems, RWTH Aachen University, Aachen, Germany

Keywords:

Agenda Planning, Information System, Intelligent Agent, Mobility as a Service, Mobility Planning, Operations

Research, Optimization Problem, Smart Sustainable Mobility, Public Transportation, Travel Planning.

Abstract:

Due to the global trend of urbanization, the current transportation networks in cities are often stressed and

congested. In the coming years, this problem is going to increase further in most areas. In addition, to

congestion of roads and public transportation, sustainability and emissions are becoming large factors when

regarding mobility. Today’s mobility still relies on the usage of private cars in a large part, however, due to the

rise of alternative travel modes and concepts such as Mobility as a Service (MaaS), the traditional mobility

market may be disrupted. With the help of MaaS and the right incentives, it may be possible to shift users

towards a more sustainable mobility behavior, which also relieve the stressed mobility network in cities, when

more alternative mobility offers are employed. With this shift towards multimodal mobility, the complexity

of searching and booking these mobility offers also rises. Users are forced to look through are large amount

of alternative offers and are required to ﬁnd a ﬁtting one. In order to make this complexity more manageable,

we propose a mobility-agenda planning agent that attends to the mobility needs of the users. The concept of

mobility-oriented agenda planning may change how people view mobility and may provide a more holistic

method of mobility planning. However, this holistic mobility planning method still has unresolved societal and

technological issues that need to be addressed.

1 INTRODUCTION

In recent years, the already complex transportation

systems in urban areas have not only grown, but also

became more heterogeneous and multimodal, further

increasing the complexity (Gallotti et al., 2016). This

development is mostly induced by two ongoing global

trends, urbanization and digitalization. The UN antici-

pates that the urbanization rate increases from today’s

per cent of the people living in urban areas to

per

cent by

2050

(United Nations, 2014). Even today, the

main roads are usually congested during peak hours

in urban areas. To handle the rising number of city

residents the already existing travel modes have to be

extended. Unfortunately, the traditional travel modes

are already at their capacity limit in certain cities, there

alternative mobility concepts, such as car- and bike-

sharing, on-demand ridesharing, have been proposed

(Cohen and Kietzmann, 2014). In Germany a shift

towards to multimodality and alternative mobility con-

cepts has already been observed (Kuhnimhof et al.,

2012). These concepts became feasible due to the on-

going efforts of digitalization in the mobility sector

(Piccinini et al., 2016). While a larger diversity in mo-

bility modes means more ﬂexibility for the passenger

on the one hand, the larger choice of transportation

modes also complicates the ﬁnding of preferred jour-

neys for the passengers on the other hand (Gallotti

et al., 2016). The main difﬁculty for the passengers

is the currently often manual combination of various

travel modes and to ﬁnd journeys that ﬁt to their prefer-

ences. One proposed solution to handle the increasing

complexity of multimodal transportation networks is

the concept of Mobility as a Service (MaaS) (Li and

Voege, 2017). MaaS promises to integrate the vari-

ous distinct mobility services into a single platform,

providing door-to-door transportation. MaaS is a com-

bination of mobility integration of existing modes, the

inclusion of new shared mobility concepts and the

usage of (mobile) applications for travel information.

For travelers the high level of integration may ease the

ﬁnding of multi-modal journeys, as the MaaS platform

is the single point of entry for mobility needs. On the

one hand, MaaS might ease the difﬁculty of using the

288

Schwinger, F. and Krempels, K.

Mobility-oriented Agenda Planning as a Value-adding Feature for Mobility as a Service.

DOI: 10.5220/0007693402880294

In Proceedings of the 11th International Conference on Agents and Artiﬁcial Intelligence (ICAART 2019), pages 288-294

ISBN: 978-989-758-350-6

 2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reser ved

inter-modal transportation system, while on the other

hand, completely new mobility services become feasi-

ble, due to a high integration of the available data from

multiple sources (Durand et al., 2018). In this paper,

we propose one of these new services that become

feasible in a MaaS scenario. While a MaaS concept

allows the user to ﬁnd a suitable mobility offer for a

single journey, it is still a burden to ﬁnd suitable mo-

bility offers for multiple journeys throughout the day.

Thus, we propose a mobility-oriented agenda planning

concept that allows the user to plan his mobility not by

planning individual trips, but by planning their agenda

by planning the appointments and tasks in a calen-

dar. The proposed agent-based system then searches

for suitable mobility offers ﬁtting to the agenda of

the users and the user’s distinct journey and trans-

portation preferences. This mobility-agenda planning

agent should reduce the cognitive load of travelers in

a multimodal urban area, when planning the mobil-

ity of their daily agenda. We attempt to tackle this

problem, by focusing on the agenda of the user and

moving the mobility itself more in the background of

the planning process. The paper is structured as fol-

lows: Section 2 further motivates the usage of MaaS,

introduces it in more detail and highlights the current

state of the art. In Section 3, we present our approach

to mobility-oriented agenda planning, whereas Sec-

tion 4 highlights the main challenges of MaaS and

mobility-oriented agenda planning and also presents

possible future work. Section 5 concludes the paper

with a brief summary.

2 STATE OF THE ART

In urban areas congestion is not the only problem with

today’s mobility. Other key factors that must be con-

sidered are, among others, sustainability and carbon

dioxide and nitrogen oxide emissions of the different

mobility modes. Due to a low capacity utilization

of private used cars, cars are often the worst way to

travel in regard to emissions (Chester and Horvath,

2009). In Germany, the average number of people

traveling in a single car was

1.5

(Follmer et al., 2010).

While commuting this factor decreases to

1.2

. That

the currently used gasoline-based cars reached their

sustainability limit can also be seen at the example of

Germany. While the usefulness of the measure is still

debated, many cities in Germany limit the roads on

which diesel-based cars may be driven to stay under

the legal NO

limit (M

ohner, 2018). Even when not

used, private cars need space in cities, which is already

sparse. When regarding commuter trafﬁc, the car not

only needs parking space at the owner’s home, but

also at the destination. Due to the aforementioned rea-

sons, individual car usage is not a sustainable mode of

transportation and alternatives for traveling are to be

preferred, in particular when regarding societal goals.

One often discussed solution to these problems is

Mobility as a Service (MaaS), which includes a mul-

titude of modes, for example car- and bikesharing or

on-demand ridesharing. In a MaaS scenario users can

either acquire their mobility on a platform, when they

need it, or buy mobility packages in advance that en-

title them to the usage of various mobility options.

MaaS offers an access-based consumption of mobility,

meaning that the mobility is based on those services

and not on the ownership of a vehicle (Bardhi and Eck-

hardt, 2012). Access-based offers are already widely

accepted in other domains, platform such as Netﬂix

and Spotify offer access to multimedia ﬁles after pay-

ing a monthly fee. MaaS attempts to apply similar

methods to the mobility market. Given the success of

platforms such as Netﬂix or Spotify, there is a huge

potential to disrupt the mobility market with MaaS.

For some people, MaaS solves all problems with to-

day’s mobility, in particular reducing the costs for all

parties involved, better management of travel demand,

and a reduction of environmental and social problems.

There is no consent on the outcome of a MaaS scenario

and its inﬂuence on the aforementioned factors. For

example, when regarding the inﬂuence of MaaS on

car usage, there are arguments for an increase and for

a decrease of car usage. While MaaS may lead to a

decrease in private car usage, because people have a

wide variety of modes to choose from on the one hand,

it also provides access to shared vehicles to people

which previously did not have access to this mode of

transportation on the other hand.

A survey study which examined numerous MaaS

studies concludes that it is especially hard to change

the travel behavior of people on their habitual trips,

particularly when no outside trigger for the people ex-

ists to change their mobility behavior (Durand et al.,

2018). The authors see the most potential for MaaS

with incidental trips at the moment. Users can be in-

cited to change their mobility behavior, for example,

by reducing the costs of their mobility or by offering

a service that adds some kind of value for the user.

Depending on the implementation of the MaaS offer,

value is already added by a larger choice of freedom.

For many people, however, the cost factor of mobility

is not as important as the convenience factor of the

mobility. The convenience of the private owned car is

the main reason for the high number of individual ve-

hicles on the roads. When looking at a MaaS scheme,

users are not subjected to a form of vendor-lock-in,

e. g. owning a yearly public transit subscription or a

Mobility-oriented Agenda Planning as a Value-adding Feature for Mobility as a Service

289

private car. Once people have accepted MaaS it may

become easier to inﬂuence their mobility behavior, as

the users do not have any invested assets for mobility

anymore, only the subscription to the MaaS service.

The users of the UbiGo MaaS prototype in Gothenburg

in Sweden, for example, adjusted their travel behavior

to a more sustainable one in a course of six months

(Karlsson et al., 2016).

In the literature, three preconditions of acceptance

are listed for MaaS: autonomy, ﬂexibility and relia-

bility (Durand et al., 2018). In this work, we want to

propose an approach for improving the reliability of

MaaS. While we do not actually make a MaaS offer

more reliable itself, we propose an agent-based system,

which, among others, watches the agenda including

the trips planned trips and informs the user of possible

breakages in the trips. Additionally, it offers to search

for alternatives. While, the reliability of the trans-

portation itself did not change, we can increase the

reliability of the overall mobility by offering alterna-

tives to the user. The ﬁrst adopters of MaaS are mostly

particularly young and tech-savvy urban individuals

(Durand et al., 2018), which may also be the audience

for a mobility-oriented agenda planning system. While

using MaaS, people also had problems with changing

between schedule-free and schedule-bound mobility.

Especially, when the mobility could not be booked as a

package. Schedule-free mobility, such as car-sharing,

still has a kind of schedule once booked, the starting

and ending time of the usage of the vehicle. As delays

in schedule-bound mobility is not uncommon, it may

negatively affect the booking window of the schedule-

free mobility. A mobility-oriented agenda agent could

also be employed to automatically adjust the bookings

of schedule-free mobility, if the service provider ac-

cepts short-term changes to a reservation, due to delays

in the previous trip. Essentially, the mobility-oriented

agenda planner, which will be introduced in the next

section is a form of adding additional convenience and

reliability to MaaS through the usage of integrated

smart services.

3 APPROACH

In order for MaaS to be successful and effective, the

user needs to perceive added value to the traditional

transportation methods and the mobility must not be

less convenient, as most people nowadays use their car

not because of time or money saved, but because of

more convenience (Durand et al., 2018). In a MaaS sce-

nario, however, with transfers between various modes,

the searching and booking of the transportation is not

convenient at all. Especially, when a multitude of

different modes are available and sensible different al-

ternative exists, it is not easy to choose an offer. Given

the access-based model of MaaS, each person has ac-

cess to all modes of mobility included in their mobility

package. To ease the usage of MaaS, we propose an

agent-based system that handles the agenda and mo-

bility needs of a user. In a well organized calendar,

appointments are entered with their respective starting

and ending time and their location. In addition to this

information, a To-Do list may also be used as input

to also partly plan the spontaneous mobility needs of

the user. We assume that the agent can learn the user’s

daily schedule over time, so that its suggestions match

the user’s requirements. One reason for this is, that the

travel behavior of people does not only repeat itself

daily, but also on a weekly, monthly or even yearly ba-

sis (Kuppam and Pendyala, 2001). In addition to that,

the agent should also learn the mobility preferences of

the user, e. g. take a shared bicycle to work, but only

when it is not raining outside.

The proposed mobility planning agent could not

only add value to the MaaS offers, but also change how

people view mobility (Wienken et al., 2017). Nowa-

days people think on a per-trip basis and book their

mobility from location A to B. The reason for the mo-

bility need, however, is nearly always an appointment

or task at location B. People rarely travel, only for the

sake of traveling itself. Even tourists mostly travel to

reach certain points of interest that are famous or that

were recommended to them. With a mobility-oriented

agenda planning the user can plan their agenda, which

many people already do anyways in the form of ap-

pointments in a calendar (Wienken et al., 2014). The

agent-based system can then look for suitable trips

that allow the user to complete their agenda on that

day. The overhead of this approach is then, that all

appointments and optional To-Dos are required to be

entered in such a way, so that the agent understand

the information. The minimal information needed for

appointments are location, starting date and time and

duration. For tasks the minimal information required

is the kind of the task, so that the agenda planner can

search for suitable locations and starting dates and

times. Additionally, further constraints such as dead-

lines, dependencies, location constraints and others

can be added by the user. The agent can then construct

an agenda from this information, which also includes

the mobility between the various locations.

An agenda planning agent, however, not only adds

beneﬁts to the user of such a system, but also to people

planning and managing the mobility of an urban area.

The mobility-oriented agenda planning approach, can

also be seen as an attempt to make spontaneous mobil-

ity planned. Mobility-oriented agendas of users can be

ICAART 2019 - 11th International Conference on Agents and Artiﬁcial Intelligence

290

used to calculate trafﬁc demands during the day. Such

data may be beneﬁcial to predict congestions in the

multimodal mobility network and offer alternatives to

the user. Offering another mode of transportation to

the user in case of a delay in another mode is called

multimodal rerouting. Allowing multimodal rerout-

ing in addition with an access-based MaaS concept

may allow to load-balance between different mobility

methods in order to better utilize the given mobility

methods in an area. This data can also be used to

improve the trafﬁc planning progress and to develop

stereotypical agendas of identiﬁed personas. Such per-

sonas can be used when simulating trafﬁc behavior of

people in cities to research the impact of either politi-

cal policy changes or the impact of new technologies

connected to mobility.

Agenda constructing problems, which are the main

task of the agent-based system, have been researched

in the domain of Operations Research. Operations

Research is a ﬁeld that uses analytical methods to

help in their decision making process. The Tourist

Trip Design Problem (TTDP), for example, deals with

constructing multiple tours in cities, which allow the

tourist to visit as many points of interest as possible,

while reducing the monetary costs or the duration of

the trips (Vansteenwegen and Van Oudheusden, 2007;

Sylejmani et al., 2017). The problem is closely related

to the Orienteering Problem or in a broader sense to

the Traveling Salesman Problem and Knapsack Prob-

lem. When regarding more general mobility-agenda

planning, additional constraints have to be taken into

account, such as deadlines, dependencies between

agenda items, or optional tasks. A ﬁrst attempt at mod-

eling the mobility-agenda planning problem builds

on the basis of the TTDP and only slightly adapted

the approach to incorporate the additional constraints.

(Schwinger et al., 2018). This solution, however, does

not compute mobility-agendas in a reasonable time,

even though insertion heuristics are employed. The

main reasons for this, is the large search space, as

many suitable locations for each task is evaluated re-

garding their insertion costs. For example, for the task

of grocery shopping, many possible supermarkets are

considered, even though the user may already has a

speciﬁc one in mind. Additionally, the agenda items

are not added one after another until reaching a local

optimum, but the complete mobility-oriented agenda

is attempted to be globally optimized, resulting in a

much larger number of calculations. This approach

was feasible for the TTDP, where a smaller number of

points of interest was regarded, but becomes unfeasi-

ble when attempting to apply the solution to a whole

country, where the number of regarded locations is

several orders of magnitude higher than when only re-

garding touristic point of interest in a smaller area. For

an agent computing mobility-oriented agendas, a long

computation time is not tolerable, as the user interacts

with the agent and expects answers in near real-time.

When limiting the search space, by involving the user

and only adding single agenda items in each query to

the agent, the computation time should be drastically

improved. The drawback of this approach is then, that

the algorithm does not converge to a global optimum,

but converges towards an local optimum and the order

in which the agenda items are added to the calendar

becomes vital.

Currently, we are examining multiple different user

interaction schemes with a mobility-agenda planning

agent. We are implementing and testing a graphical

user interface (GUI), a voice user interface (VUI) and

a text-based natural language processing (NLP) fron-

tend. All applications have access to a multimodal

route planner and a mockup calendar of a user. The

user interactions are similar in each scenario, only the

input and output paths from and to the user are dif-

ferent. The user is able to add, remove and change

appointments and To-Dos. Furthermore, the mobility

preferences and other constraints can be modiﬁed. Mo-

bility preferences includes the mode preference, the

variables that should be minimized, e. g. traveling time,

number of transfers or monetary costs. The GUI, for

example, features an calendar, a To-Do list view and

a map view, showing the mobility between the differ-

ent locations of the agenda. In the GUI the complete

mobility-oriented agenda is visualized, normally users

are only able to see the appointments in a calendar and

need to look up traveling times themselves using a trip

planning service.

The main feature of the agent, while also crucial,

is not the planning aspect of the agenda. The agent

should be able to deal with disruptions to the agenda

and attempt to offer solutions to the user when certain

aspects of the agenda are not feasible anymore. This

includes problems with the mobility itself, e. g. the

train is delayed and the next appointment cannot be

reached by train in time. The agent could then recom-

mend to book a shared-car that allows the user to still

reach their appointment on time. Since the agent reacts

to real-world events, the agent needs a real-time feed

of mobility information. This includes delays of pub-

lic transportation vehicles, current and prognosticated

trafﬁc information and booking availability of shared

vehicles. Other problems to the agenda may occur,

when the user does not follow the planned agenda, for

example, due to a meeting that takes longer than ini-

tially anticipated. The agenda planning agent may sug-

gest, for example, to cancel the following appointment

or to move tasks that were planned for that speciﬁc

Mobility-oriented Agenda Planning as a Value-adding Feature for Mobility as a Service

291

time slot to a later time. In our proposed solution the

agenda planning agent never acts on its own, but only

suggests possible actions to the user. The human is

always in the loop and can also choose to not follow

the proposed actions.

4 OPEN CHALLENGES

When tackling mobility-oriented agenda planning,

there are a few challenges that need to be addressed.

We grouped these challenges in two coarse groups,

social challenges and technological challenges.

Social Aspects.

The social challenges address prob-

lems that arise when attempting to deploy a mobility-

oriented agenda agent. These challenges are required

to be addressed to avoid that people either do not want

use newly developed system or that they are unable to

use it.

Setup.

One challenge with the approach of mobility-

agenda planning in general is that its usefulness

only fully unfolds when combining it with a MaaS

scheme. When all trips are completed per car, the

added value of a mobility-oriented agenda planner

is negligible for most users. Therefore, the concept

of mobility-oriented agenda planning is closely

connected to implemented MaaS schemes. These

offers, however, are not widely available, and the

available prototypes are focused on speciﬁc urban

areas. Furthermore, MaaS offers are only expected

to come to urban areas in the coming years.

Mobility Setup.

The current research focuses mostly

on the potential early-adopters of a MaaS scheme,

mostly categorized as urban living tech-savvy

young people. Rural areas are, mostly due to cost

considerations, left out (Durand et al., 2018). The

assumed backbone of the MaaS scheme, public

transportation (Kamargianni et al., 2018), is also

not widely available in rural areas. In rural areas,

schedule-based mobility such as public transporta-

tion is very infrequent and therefore no alternative

to owning a car (Shergold et al., 2012).

Digital Divide.

Additionally, MaaS and also mobility-

oriented agenda planning requires the use of digital

technologies (Jittrapirom et al., 2017). It has to

be ensured, however, that no digital divide occurs

and also elderly people can use these new mobility

offers. Especially, elderly people, who do not own

a car could beneﬁt from the adoption of MaaS

(Stein et al., 2017).

Habitual Behavior.

Another problem that needs to

be addressed by a mobility-oriented agenda plan-

ner is that people often prefer the status quo, in-

stead of trying out new things (Ratilainen, 2017).

When people book their MaaS package and do not

change their mobility habits, the goal of a more

sustainable way of traveling is not reached. There-

fore, triggers that encourage people to try out new

mobility modes must be found. One method would

be to incorporate trialability into the offers, so that

the user can try other mobility modes for a subsi-

dized price. A gamiﬁcation aspect, where users

are competing for reducing their carbon footprint,

could also be added (Kazhamiakin et al., 2015).

Privacy.

Another problems with the concept of

mobility-oriented agenda planning lies with the pri-

vacy aspect. In order to plan the mobility-oriented

agenda and offering the user tailor-made mobility,

the agent needs the complete calendar information

including the preferences of the user. Most people

are, however, not willing to share this sensitive data

with an external service provider. This is still an

unsolved challenge and is required to be addressed

in the future and is a crucial question.

Most of the social aspects affect the MaaS concept

itself. The main social challenge an approach of the

mobility-oriented agenda planner must solve, is the

aspect of privacy.

Technological Aspects.

The technological chal-

lenges highlight the main problems the developer of

such a system has to resolve. Some of these challenges

are functional, while others are non-functional.

Multimodality.

To compute meaningful trips and

agendas the optimization algorithm must be able

to access a multimodal multi-objective router with

real-time information. The multimodality aspect

is required to offer multiple different trips to the

users, these trips themselves may consist of multi-

ple modes. To allow the user to specify their own

preferences about their mobility trips, the router

must be able to handle multiple objectives. The

user may want to minimize the travel time, the

monetary costs, or the amount of vehicle inter-

changes. Unfortunately, just combining the dif-

ferent objectives using a linear combination and

weighted factors may lead to deteriorated results

(Delling et al., 2012).

Actuality.

As the user interacts with the agent, the

various computations of the agent should not take

too much time. One non-functional challenge is,

to implement the scheduling algorithm and the

underlying routing algorithm in such a way that

a normal use of the system is possible. Finally,

real-time information is needed to address current

ICAART 2019 - 11th International Conference on Agents and Artiﬁcial Intelligence

292

congestion on the roads or delays in the public

transit network. The mobility-agenda planner then

knows the current predictions about the travel time

from one location to another to compute the agenda

in the ﬁrst place, but also to update the agenda

in case of expected delays in some modes. This

leads to the next challenge with mobility-oriented

agenda planning.

Foresight.

Any time unforeseen events may happen

during the day and affect the already planned

agenda. Certain appointments or tasks may take

longer than anticipated, or the trip between two

locations is prolonged due to congestion. On the

one hand, the agent must continuously monitor

the agenda to inform the user of certain agenda-

breaking events, on the other hand, the user must

also be able to inform the agent to changes to the

agenda. Once a change to the agenda is communi-

cated by either side, the mobility agenda must be

adapted to become feasible again.

Individuality.

Another challenge that needs to be ad-

dressed is that the agenda of a user is very per-

sonal. The preferences regarding mobility, the

location of agenda items, or the time and order

of agenda items strongly diverges between differ-

ent users. One possible solution to this is to learn

the user’s preferences regarding time and location

of appointments and the user’s choice of mobility

for certain agenda items. Additionally routines of

the user can be learned, for example, when and

where the user prefers to go shopping. This is a

promising approach, as the mobility habits of users

tend to repeat themselves (Kuppam and Pendyala,

2001). Various learning methods have been pro-

posed in the literature for recommendations re-

garding agenda items (Oh and Smith, 2004) or for

mobility (Arentze, 2013). For both problems, the

user input to the agent can be used as a basis for the

learning. Firstly, the user may already explicitly

state their preferences when addressing the agent.

Secondly, upon returning a result to the user, she

is able to manually change parts of the mobility

and/or agenda. In this case, manual changes to the

solution correspond to non-met user preferences.

In both scenarios, the agent can learn from the user

interaction in order to improve its understanding

of the user’s preferences.

The technological aspects can be grouped into three

groups. Speed of the computations, quality of the

results and quality of the interaction with the agent.

Currently our focus is on delivering quality results and

designing interaction schemes with the agent.

5 CONCLUSION

In this paper, we have discussed problems with today’s

mobility, especially regarding their efﬁciency and sus-

tainability. Given the recent trends, new mobility con-

cepts are required under both aspects. Current trans-

portation methods are at their limit, regarding their

throughput and their sustainability. As one promis-

ing solution to this problem, we have introduced the

concept of Mobility as a Service. We explained the

reasons why a MaaS concept may better utilize the

already available mobility methods and why MaaS,

with the right incentives, may lead to more sustainable

mobility behavior of the users. Furthermore, the con-

cept of mobility-oriented agenda planning has been

introduced as a way to add value to MaaS in form of

a smart service. For this we have illustrated an agent-

based system that helps users to ﬁnd suitable mobility

for their agenda and explained the drawbacks of the

traditional trip-based mobility search paradigm. In the

mobility-oriented agenda planning concept, the user

does not plan their mobility directly, but the mobil-

ity is rather seen as a way to reach the locations of

the various agenda items, appointments and To-Dos.

As the agent knows the agenda for the complete day,

more reasonable suggestions may be given to the users,

compared to a trip-based itinerary planner. Finally, we

highlighted the current social and technological chal-

lenges with mobility-oriented agenda planning and

have shown where future work is required.

For future work, further scenarios for a mobility-

oriented agenda planning system could be developed

to derive additional requirements of the system. With

the requirements the architecture and system design

can also be reﬁned. The open social and technological

challenges should also be addressed in future work.

REFERENCES

Arentze, T. A. (2013). Adaptive personalized travel infor-

mation systems: A bayesian method to learn users’

personal preferences in multimodal transport networks.

IEEE Transactions on Intelligent Transportation Sys-

tems, 14(4):1957–1966.

Bardhi, F. and Eckhardt, G. M. (2012). Access-based con-

sumption: The case of car sharing. Journal of con-

sumer research, 39(4):881–898.

Chester, M. V. and Horvath, A. (2009). Environmental as-

sessment of passenger transportation should include

infrastructure and supply chains. Environmental re-

search letters, 4(2):024008.

Cohen, B. and Kietzmann, J. (2014). Ride on! mobility busi-

ness models for the sharing economy. Organization &

Environment, 27(3):279–296.

Mobility-oriented Agenda Planning as a Value-adding Feature for Mobility as a Service

293

Delling, D., Dibbelt, J., Pajor, T., Wagner, D., and Werneck,

R. F. (2012). Computing and evaluating multimodal

journeys. KIT, Fakult

at f

ur Informatik.

Durand, A., Harms, L., Hoogendoorn-Lanser, S., and Zi-

jlstra, T. (2018). Mobility-as-a-service and changes

in travel preferences and travel behaviour: a literature

review.

Follmer, R., Gruschwitz, D., Jesske, B., Quandt, S., Lenz,

B., Nobis, C., K

ohler, K., and Mehlin, M. (2010).

Mobilit

at in deutschland 2008: Struktur–aufkommen–

emissionen–trends. Bonn: infas Institut f

ur ange-

wandte Sozialwissenschaft GmbH.

Gallotti, R., Porter, M. A., and Barthelemy, M. (2016). Lost

in transportation: Information measures and cogni-

tive limits in multilayer navigation. Science advances,

2(2):e1500445.

Jittrapirom, P., Caiati, V., Feneri, A.-M., Ebrahimighare-

hbaghi, S., Gonz

alez, M. J. A., and Narayan, J. (2017).

Mobility as a service: a critical review of deﬁnitions,

assessments of schemes, and key challenges. Urban

Planning, 2(2):13.

Kamargianni, M., Matyas, M., Li, W., and Muscat, J.

(2018). Londoners’ attitudes towards car-ownership

and mobility-as-a-service: Impact assessment and op-

portunities that lie ahead.

Karlsson, I. M., Sochor, J., and Str

omberg, H. (2016). Devel-

oping the ‘service’in mobility as a service: experiences

from a ﬁeld trial of an innovative travel brokerage.

Transportation Research Procedia, 14:3265–3273.

Kazhamiakin, R., Marconi, A., Perillo, M., Pistore, M.,

Valetto, G., Piras, L., Avesani, F., and Perri, N. (2015).

Using gamiﬁcation to incentivize sustainable urban

mobility. In Smart Cities Conference (ISC2), 2015

IEEE First International, pages 1–6. IEEE.

Kuhnimhof, T., Buehler, R., Wirtz, M., and Kalinowska, D.

(2012). Travel trends among young adults in germany:

increasing multimodality and declining car use for men.

Journal of Transport Geography, 24:443–450.

Kuppam, A. R. and Pendyala, R. M. (2001). A structural

equations analysis of commuters’ activity and travel

patterns. Transportation, 28(1):33–54.

Li, Y. and Voege, T. (2017). Mobility as a service (maas):

Challenges of implementation and policy required.

Journal of transportation technologies, 7(02):95.

ohner, M. (2018). Driving ban for diesel-powered vehicles

in major cities: an appropriate penalty for exceeding

the limit value for nitrogen dioxide? International

Archives of Occupational and Environmental Health,

4(91):373–376.

Oh, J. and Smith, S. F. (2004). Learning user preferences

in distributed calendar scheduling. In International

Conference on the Practice and Theory of Automated

Timetabling, pages 3–16. Springer.

Piccinini, E., Flores, C. K., Vieira, D., and Kolbe, L. M.

(2016). The future of personal urban mobility–towards

digital transformation. Wirtschaftsinformatik (MKWI),

page 55.

Ratilainen, H. (2017). Mobility-as-a-service: Exploring

consumer preferences for maas subscription packages

using a stated choice experiment.

Schwinger, F., Ohler, F., and Krempels, K.-H. (2018). Flex-

ible multicriterial agenda planning in public transit

systems - an intelligent agent for mobility-oriented

agenda planning. In Proceedings of the 4th Interna-

tional Conference on Vehicle Technology and Intelli-

gent Transport Systems - Volume 1: VEHITS, pages

558–567. INSTICC, SciTePress.

Shergold, I., Parkhurst, G., and Musselwhite, C. (2012). Ru-

ral car dependence: an emerging barrier to community

activity for older people. Transportation Planning and

Technology, 35(1):69–85.

Stein, M., Meurer, J., Boden, A., and Wulf, V. (2017). Mo-

bility in later life: Appropriation of an integrated trans-

portation platform. In Proceedings of the 2017 CHI

Conference on Human Factors in Computing Systems,

pages 5716–5729. ACM.

Sylejmani, K., Dorn, J., and Musliu, N. (2017). Planning the

trip itinerary for tourist groups. Information Technol-

ogy & Tourism, 17(3):275–314.

United Nations (2014). World urbanization prospects: The

2014 revision, highlights. department of economic and

social affairs. Population Division, United Nations.

Vansteenwegen, P. and Van Oudheusden, D. (2007). The

mobile tourist guide: an or opportunity. OR insight,

20(3):21–27.

Wienken, T., Kr

omker, H., and Spundﬂasch, S. (2017).

Agenda planning-design guidelines for holistic mobil-

ity planning. In International Conference on Human-

Computer Interaction, pages 713–720. Springer.

Wienken, T., Mayas, C., H

orold, S., and Kr

omker, H. (2014).

Model of mobility oriented agenda planning. In Inter-

national Conference on Human-Computer Interaction,

pages 537–544. Springer.

ICAART 2019 - 11th International Conference on Agents and Artiﬁcial Intelligence

294