=Paper=
{{Paper
|id=None
|storemode=property
|title=Peacox - Persuasive Advisor for CO2-Reducing Cross-Modal Trip Planning
|pdfUrl=https://ceur-ws.org/Vol-973/bcss5.pdf
|volume=Vol-973
|dblpUrl=https://dblp.org/rec/conf/persuasive/SchrammelBT13
}}
==Peacox - Persuasive Advisor for CO2-Reducing Cross-Modal Trip Planning
==
https://ceur-ws.org/Vol-973/bcss5.pdf
First International Workshop on Behavior Change Support Systems (BCSS 2013) 29
Peacox – Persuasive Advisor for CO2-Reducing
Cross-modal Trip Planning
Johann Schrammel1, Marc Busch1 and Manfred Tscheligi1,2
1
CURE – Center for Usability Research & Engineering, Vienna, Austria
(schrammel,busch,tscheligi)@cure.at
2
University of Salzburg, ICT&S Center, Salzburg, Austria
manfred.tschligi@sbg.ac.at
Abstract. We present the objectives and first results of the Peacox project. Pea-
cox aims at providing travelers with personalized multi-modal navigation tools
that allow, help and persuade them to travel and drive ecological friendlier. To
support the users in making travel decisions without feeling restricted, Peacox
considers their situational and individual range of acceptable travel choices and
provides personalized and tailored navigation support.
1 Project Overview, Motivation and Objectives
PEACOX - Persuasive Advisor for CO2-reducing cross-modal trip planning - is an
international research project aiming to provide travelers with personalized multi-
modal navigation tools that allow, help and persuade them to travel and drive ecologi-
cal friendlier. Existing navigation tools typically focus either on one trip mode, do not
offer personalization or do not stress the ecological aspect. As ecological issues in
environments become more and more pressing, means to reduce the ecological impact
of people are needed urgently. A substantial part of emissions is related to traffic and
mobility, and as work and leisure life becomes more and more geographically distrib-
uted it has become even more important to support and guide users to behave pro-
environmental with regard to their travelling behavior and decisions.
To answer this need the main approach of Peacox is to provide travelers with per-
sonalized mobile and web tools that allow, help and persuade them to plan their trip
aiming at the environmentally friendliest travel modes. To support the users in mak-
ing this decision without feeling restricted, Peacox will consider their situational and
individual range of acceptable travel choices. Peacox researches possibilities for in-
fluencing travel and driving behavior of users with the use of targeted persuasive
strategies, providing situated and personalized feedback and use of advanced travel
information systems.
Within Peacox, we develop mobile and web applications that enable the users on
the move and at home to easily plan and organize their trip (by foot, bike, public
transport, motorcycle, car, and by use of car pooling).
To achieve maximum impact, Peacox provides situated and located suggestions
and information to the users regarding travel choices and options. The calculation and
30 First International Workshop on Behavior Change Support Systems (BCSS 2013)
presentation of options considers the current location of the users, their actual travel
situation, their individual preferences as well as their travel mode choice and trip
history. Considering all the relevant context information, the system aims at present-
ing the user only attractive and relevant suggestions.
Peacox reduces the need for explicit inputs by the user and thereby aims at increas-
ing the user experience, comfort and willingness to use. This expected effect is a main
research goal for the field evaluations of the system. In contrast to existing trip plan-
ning and impact calculation services, Peacox automatically keeps track of the users
prior travel decisions and tracks (by use of GPS and automated travel mode detec-
tion), identifies the current mode and purpose of a trip, and builds tailored models for
each user. The user therefore is not required to e.g. always specify which means of
transportation she/he is using to receive proper recommendations.
Peacox calculates the ecological/carbon footprint considering the used means of
transportation as well as dynamic variables influencing the actual emissions, such as
current traffic situation and therefore can provide more accurate data than simplistic
and static computation models. Besides CO2, the model also considers additional
emissions such as NOX, SO2 and PM10.
Peacox further presents relevant information for multi-modal trip planning and en-
vironmental impact feedback using persuasive interfaces targeted at reinforcing desir-
able behavior. The systems interfaces are designed with special consideration of guid-
ing the user to choose less polluting alternatives utilizing known psychological prin-
ciples and strategies, e.g. by making the consequences of choices clearly visible dur-
ing the decision process. Special care is given during the design to focus on emotion-
ally positive aspects rather than restrictive approaches. The current design incorpo-
rates the following persuasive strategies: Suggestion, Tunnelling, Conditioning, Self-
Monitoring [4], as well as Social Proof, Consistency and Authority [3].
2 System Design
Figure 1 below depicts the core components of the Peacox-System and their inter-
play. The image shows a distinction between the frontend system (on the top left side
of the visualization) and the backend system. In the frontend the user specifies his/her
destination (Peacox provides already destination suggestions based on trip history)
and the system then calculates and displays recommendations of different travel mode
alternatives.
These recommendations are based on a process in the backend system. With the
aid of GPS dynamically updated location data is collected from the user. After identi-
fying trip segments and points of interest (POIs), Peacox classifies this data and sub-
sequently creates a trip history for each individual user. With this information about
previous travel behavior Peacox identifies travel patterns e.g. one user always uses
public transport on the way to work. Those travel patterns include information about
traffic mode, time and duration, preferences and location. Peacox uses this data to
personalize the recommendations for travel behavior to the individual users, so users
First International Workshop on Behavior Change Support Systems (BCSS 2013) 31
are only presented recommendations that are within reasonable limits for the given
user.
For each trip Peacox models the individual ecological footprint and calculates trip
alternatives and their related ecological costs. This information allows to evaluate the
environmental impact of different trip alternatives, which is then used in the user
interface to influence the decision of the users by the mentioned persuasive strategies.
Furthermore, information about actual impact of a user travels and possible savings is
collected to be used for retrospective analysis.
Trip History, Travel Pattern Identification, modeling the environmental footprint
and calculating trip alternatives get information values from GPS, weather, real-time
traffic and statistical data. To increase accuracy and privacy the user has the possibil-
ity (but is not required) to verify the outcome of the detection process and to specify
preferences and privacy settings which helps to increase the helpfulness of the system.
Fig. 1. Peacox System Concept
3 Personalized and Situated Persuasion
As outlined above one key concept of Peacox is to use personalized and situated per-
suasive strategies. Our main approach to the development of a system in accordance
with this goal is to carefully balance and orchestrate the properties of persuasive strat-
egies with the needs arising from contextual/situational variables and the characteris-
tics of different user groups and personality types.
In order to do so on an implementation level we analyzed and classified persuasive
strategies with regard to their suitability for different application scenarios, developed
a classification and prioritization of context variables influencing mobility decisions
and identified and analyzed user groups from two different perspectives: First, we had
a close look on different trip choice models and typologies [e.g. 1, 5] and second we
32 First International Workshop on Behavior Change Support Systems (BCSS 2013)
studied the influence of the travelers environmental attitudes (measured by use of
standardized inventories such as e.g. [8] ) on trip mode choice.
To create personalized persuasive technology (which is expected to have greater
impact than not-personalized technology), and to tailor the persuasive approach to the
individual user it is also necessary to be able to estimate the susceptibility of a person
to different persuasive strategies (persuadability) - this is also referred to as “Persua-
sion Profiling” [6]. We use available inventories [7] and also developed and validated
our own methods [2] that can be used to estimate susceptibility to persuasive strate-
gies to personalize persuasive technology according to the users’ personality based on
self-reports. In future work we aim at deducing persuadability characteristics by ana-
lyzing the users’ behavior and reaction following persuasive interventions, thereby
removing the need for explicit input from the user.
A second important aspect for Peacox is the tailoring of persuasive interventions
based on understanding of the users travel context and preferences. This understand-
ing is based on automated trip mode detection and trip purpose imputation. Trip mode
is analyzed using GPS and map data, and works already sufficiently. In an example
trial accuracy of around 83% could be achieved. Trip purpose imputation uses activity
and location (both person related and general) as input parameters to detect basic
types such as home, work/education or shopping. Both, trip mode and trip purpose are
used to tailor the persuasion to the current situation and users need.
We currently are working on developing an overall model that integrates the dif-
ferent aspects influencing the persuasive approach and content. Integration is done
both, based on theoretical considerations as well as based on empirical data collected
by different means (questionnaires, experiments and field studies). Furthermore, de-
sign concepts on how the different elements can be best integrated with basic naviga-
tion functionality have been developed and are currently evaluated with end users.
References
1. Anable, J. (2005): „Complacent Car Addicts‟ or „Aspiring Environmentalists‟? Identify-
ing travel behavior segments using attitude theory. Transport Policy, 12, 65–78.
2. Busch, M., Schrammel, J. and Tscheligi, M. Personalized Persuasive Technology – Devel-
opment and Validation of Scales for Measuring Persuadability. 8th International Confer-
ence on Persuasive Technology, Sydney, Australia, 2013.
3. Cialdini, R.B. (2001): Harnessing the Science of Persuasion. Harvard Business Review.
4. Fogg, B. J. (2003): Persuasive Technology. Using Computers to Change What We Think
and Do. San Francisco, Elsevier, Page 31-53.
5. Hunecke, M. et al. (2008): Attitude-Based Target Groups to Reduce the Ecological Impact
of Daily Mobility Behavior. Environment and Behavior.
6. Kaptein, M., Markopoulos, P.: Can you be persuaded? individual differences in
susceptibility to persuasion. INTERACT (2009).
7. Kaptein, M.C.: Personalized persuasion in Ambient Intelligence. Journal of Ambient
Intelligence and Smart Environments. (2012).
8. Milfont, T.L. & Duckitt, J. (2010): The environmental attitudes inventory: A valid and re-
liable measure to assess the structure of environmental attitudes. Journal of Environmental
Psychology. 30, 80-94.