This is the author’s version of a work that was submitted/accepted for pub-
lication in the following source:

Rakotonirainy, Andry, Schroeter, Ronald, & Soro, Alessandro
(2014)
Three social car visions to improve driver behaviour.
Pervasive and Mobile Computing, 14, pp. 147-160.

This file was downloaded from: https://eprints.qut.edu.au/73042/

c© Copyright 2014 Please consult the authors

NOTICE: this is the author’s version of a work that was accepted for publication in Pervasive
and Mobile Computing. Changes resulting from the publishing process, such as peer
review, editing, corrections, structural formatting, and other quality control mechanisms
may not be reflected in this document. Changes may have been made to this work since it
was submitted for publication. A definitive version was subsequently published in Pervasive
and Mobile Computing, [VOL#, ISSUE#, (DATE)] DOI: 10.1016/j.pmcj.2014.06.004

Notice: Changes introduced as a result of publishing processes such as
copy-editing and formatting may not be reflected in this document. For a
definitive version of this work, please refer to the published source:

https://doi.org/10.1016/j.pmcj.2014.06.004

https://eprints.qut.edu.au/view/person/Rakotonirainy,_Andry.html
https://eprints.qut.edu.au/view/person/Schroeter,_Ronald.html
https://eprints.qut.edu.au/view/person/Soro,_Alessandro.html
https://eprints.qut.edu.au/73042/
https://doi.org/10.1016/j.pmcj.2014.06.004


Three Social Car Visions to Improve Driver Behaviour

Andry Rakotonirainya,, Ronald Schroetera, Alessandro Soroa,b

aCentre for Accident Research Road Safety - Queensland
bComputer Human Interaction Group
Queensland University of Technology

Australia

Abstract

The social cost of road injury and fatalities is still unacceptable. The driver
is often the main responsible for road crashes, therefore changing the driver
behaviour is one of the most important and most challenging priority in road
transport. This paper presents three innovative visions that articulate the
potential of using Vehicle to Vehicle (V2V) communication for supporting
the exchange of social information amongst drivers. We argue that there
could be tremendous benefits in socializing cars to influence human driving
behaviours for the better and that this aspect is still relevant in the age of
looming autonomous cars. Our visions provide theoretical grounding how
V2V infrastructure and emerging human machine interfaces (HMI) could
persuade drivers to (i) adopt better (e.g. greener) driving practices, (ii)
reduce drivers aggressiveness towards pro-social driving behaviours, and (iii)
reduce risk-taking behaviour in young, particularly male, adults. The visions
present simple but powerful concepts that reveal ‘good’ aspects of the driver
behaviour to other drivers and make them contagious. The use of self-efficacy,
social norms, gamification theories and social cues could then increase the
likelihood of a widespread adoption of such ‘good’ driving behaviours.

Keywords: social cars, pervasive computing, social norms, cooperative
systems, Intelligent Transport Systems

Email addresses: r.andry@qut.edu.au (Andry Rakotonirainy),
r.schroeter@qut.edu.au (Ronald Schroeter), alessandro.soro@qut.edu.au
(Alessandro Soro)

Preprint submitted to PERVASIVE AND MOBILE COMPUTING July 1, 2014



1. Introduction

Human’s craving for social connectedness continues to grow. Social net-
works have used web technology to fulfil our insatiable need for social con-
nectedness and has unexpectedly changed the way individuals interact with
each other. Smartphone devices have brought social networks into our cars,
however, due to the vehicle’s safety requirements, social networks have never
been fully integrated into our driving environment. Furthermore, the physical
nature of vehicles such as the metal shield prevents natural social interactions
between road users. Cooperative systems such as Vehicle to Vehicle (V2V)
communication offer new location aware services which will allow drivers, in
the same vicinity, to share and exchange situational information anywhere
and anytime. This, however, raises the elementary questions proposed by
this special issue, including ‘who can communicate what, when, how, and
why?’

The aim of this paper is to present three innovative visions that furthers
the debate around those questions, in particular the question ‘why’. We
argue that social pressure are particularly suitable to influence human driving
behaviours for the better and that this aspect is still relevant in the age of
looming autonomous cars. Our visions provide theoretical grounding how
V2V infrastructure and emerging human machine interfaces (HMI) could
persuade drivers to (i) adopt ‘better’ (e.g. greener) driving practices, (ii)
reduce drivers aggressiveness towards pro-social driving behaviours, and (iii)
reduce risk-taking behaviour in young, particularly male, adults.

1.1. Why: Humans still matter

Improving transportation efficiency through operational innovation is crit-
ical as our population grows and ages, budgets tighten and consumer prefer-
ences shift. Aside from important issues such as reducing road trauma, fuel
consumption and emissions, we also need future technologies to accommo-
date for the road users social needs. The ‘Google autonomous car’ is claimed
to be ready for the public within 5 years whilst Volvo is predicting that within
seven years “you won’t be able to crash” its cars by using semi-autonomous
technologies such as low-speed collision-avoidance or pedestrian detection.

Although these valuable endeavours will make significant contributions
to transport and safety by removing the control from the driver; our visions
complements the (semi-) autonomous technologies where the driver is still
responsible for operating the vehicle at strategic, tactical and operational

2



levels [62]. It has been shown that driver’s decision making is heavily influ-
enced by the social setting. Driving is a socially regulated behaviour [91].
Normative factors have been shown to have influence on speeding behaviour;
for example, how driver perceptions of the beliefs, attitudes, and actions of
their peers towards speeding can influence driving speeds [29].

The following subsections draw attention to the question as to ‘how’ social
information will be able to be communicated in real-time. They introduce
several underlying, emerging technologies that will ultimately enable our
visions.

1.2. How: Intelligent Transport Systems and Cooperative Systems

Intelligent Transport Systems (ITS) concern the use of information and
communication technologies applied to transport infrastructure and vehicles.
ITS have the potential to reduce fatalities and injuries by 40% across the
OECD [65]. Existing autonomous Advanced Driving Assistance Systems
(ADAS), which are a subset of ITS, include examples such as pedestrian
detection in bad vision conditions. They generally use various sensors such
as radars, cameras, or lasers to gather contextual/situational information
about the vehicle’s surroundings (e.g., pedestrian approaching) in order to
provide appropriate countermeasures (e.g., warning or breaking). However,
such systems often have technical limitations. For example, the sensors’
perception and awareness are limited to the immediate surrounding area of
the vehicle, and this can be obscured.

Cooperative systems, which allow vehicles to communicate with each
other to achieve a common goal, are widely recognised as the next big chal-
lenge in ITS (http://www.cvisproject.org). Cooperative systems can offer
significant improvements in the safety of all road users by increasing drivers’
awareness given that 80% of vehicle crashes are due to human errors. Co-
operative systems can also improve the quality and reliability of information
available to drivers about their immediate and distant environment.

Most of existing approaches consist of exchanging the current vehicle
kinematics and the whereabouts of hazards between two vehicles with the
view to anticipate crash avoidance. However, cooperative systems such as
V2V could be used to facilitate location aware peer to peer communications
between road users.

ITS resembles the infrastructure for ubiquitous computing in the car. It
encompasses a) all kinds of sensing technologies within vehicles as well as

3



road infrastructure, b) wireless communication protocols for the sensed in-
formation to be exchanged between vehicles (V2V) and between vehicles and
infrastructure (V2I), and c) appropriate intelligent algorithms and computa-
tional technologies that process these real-time streams of information. As
such, ITS can be considered a game changer. It provides the fundamental
basis of new, innovative concepts and applications, similar to the Internet
itself.

The information sensed or gathered within or around the vehicle has led
to a variety of context-aware in-vehicular technologies within the car. A sim-
ple example is object detection, which stops the vehicle when sensors (camera
or radar) detect an object within the the trajectory. We refer to this type
of context awareness as vehicle/technology awareness. V2V and V2I com-
munication, often summarised as V2X, enables the exchange and sharing of
sensed information amongst cars. As a result, the vehicle/technology aware-
ness horizon of each individual car is expanded beyond its observable sur-
rounding, paving the way to technologically enhance such already advanced
systems.

1.2.1. Limitation: Security

Cooperative systems will exacerbate the need for strong security. Driving
performance, location and identity (car registration) could easily be tracked
with existing technology. Privacy has been hailed as a potential major issue
in cooperative systems. The human user is often the weakest link in the
security chain of a software system. Changing security profile whilst driving
is cognitively more challenging than performing it in a desktop environment.
A poor security usability in the V2X context could lead to serious security
vulnerabilities that can be exploited for criminal purpose. “The system must
be easy to use and must neither require stress of mind nor the knowledge of
a long series of rules” [48].

1.2.2. Limitation: Reliability of exchanged information

The Human Computer Interfaces (HCI) research community has been
looking at how drivers could most effectively receive and act upon received
information and how information can be delivered with the least amount of
unintended consequences, such as distraction [74]. The design of HCI for
cooperative applications is still in its infancy. Most in-vehicle HCI research
was conducted for Advanced Driving Assistance Systems (ADAS) with the
underline hypothesis that the information presented to the driver are highly

4



accurate and reliable. Such assumptions are no longer valid in cooperative
systems environments using vehicular ad-hoc networks (VANET). In previ-
ous work we have shown that the 802.11 broadcasts service, which is the
building block of V2V communications, cannot be reliable [21]. Each car
can have different perception of their location (e.g. by using GPS), reaching
agreement between different cars about their respective relative positions re-
quires complex data fusion and conflict resolution mechanisms which do not
necessarily provide the most accurate information. In [37] we further explore
the possibility of modeling and simulating the benefits of Cooperative Sys-
tems based on Inter-Vehicular Communication with the aim of implementing
a freeway emergency braking scenario. While our model confirms (together
with related research) that collaborative system can reduce the number of
crashes, yet the average crash severity appears to remain constant, which
stresses on the urgency of more work on such topics.

1.3. How: New HCI & augmented reality

Novel HCI paradigms, such as tangible, gestural and manipulative in-
terfaces, draw their theoretical soundness from various theories of embodied
cognition, e.g. Kirsh and Maglio’s epistemic action [49, 56], Dourish’s em-
bodied interaction [23] and Hostetter and Alibali’s visible embodiment [42].
While these works have had a tremendous impact on human-computer in-
teraction (HCI) and particularly on Computer Supported Cooperative Work
(CSCW), a translation to the road transport domain is necessary to take into
account the peculiarities of the interaction between drivers. In addition, any
theory or model of human machine interaction in the context of driving be-
haviour that fails to include social concepts could be argued to lack a critical
element.

Our visions are placed in future scenarios where we assume novel inter-
faces to be context aware and natural to effectively convey information when
it is safe to do so, i.e., when they do not increase the information overload
of the driver.

Conceptually, our visions are irrespective of the modality of interaction,
although the examples in the following sections are depicted visually. Nev-
ertheless, given the current state of AR glasses and HUDs, we do see the
biggest potential in visual interfaces to convey (so output rather than input)
information in such way that it augments reality with an increased ambient,
social awareness. We therefore propose to introduce these augmented reality
concepts via Head Up Displays (HUDs).

5



In the past, HUDs have been developed and primarily used in fighter jets
as a way to convey visual information to the pilots in the least distracting way.
In luxury cars, HUDs are considered to be the safest way to convey visual
information to the driver, e.g., speed, navigation, etc. In the not so distant
future, such displays are going to enter a wider consumer market, examples
include new wearable computing devices, such as Augmented Reality (AR)
glasses (e.g. Google Project Glass), or HUDs specifically designed to be
retrofitted in cars (e.g. Pioneer CyberNavi).

Using such technologies to convey social information in cars is a relatively
new field. The HCI can augment the driver’s social perception by adding,
in his/her field of view, elements that improve social awareness, essentially
creating what we refer to as the ‘social car’ [82] [83]. We have shown that
the use of, for example, avatars on HUDs can help drivers to safely negotiate
intersections [72].

1.4. Social car

As mentioned above, in this paper, we draw attention to those application
areas of sensing and V2X technologies, where the driver’s behavior and hence
the socio-psychological perspective plays a more pivotal role. The focal points
of our research around the social car is illustrated in Fig. 1. The vehicle first
(1) gathers or senses social information about the driver and the driving
behaviour. We purposely keep the definition of what that information might
be very abstract, but will provide concrete examples in our vision sections.
Generally, this information can range from broad social network information
to the driver’s current facial expression, or from the driver’s historical driving
data to the current revs of the engine.

Note, that it is not the type of information that makes it social, but the
fact that it is shared. For example, applications that gather or sense infor-
mation and feed it back to the driver (2 in Fig. 1) does create an increased
self-awareness in itself (a simple example is the speedometer). However, as
we will show in vision A, this self-awareness has the potential to be intensified
through the social layers.

Hence, our definition of the social car is a car with the ability to share
social information with other road users through V2X technologies, e.g. ve-
hicular ad-hoc networks (VANET) [81]. Using these V2X technologies, the
sensed information can then be (3) passed to surrounding drivers for an in-
creased direct social awareness, or (4), pushed even further, into the cloud,

6



where it is collected and visualized for an increased, collective social aware-
ness within the driving community at large. Our visions presented in this
paper focus on these levels of social awareness with a view to improve driving
behaviours.

2. Vision A: Social norms and self-efficacy to motivate better driv-
ing behaviours

The development of future in-car technology interventions can borrow
design strategies from outside the driving context. Previous technologies,
for example, have aimed at helping users to change behaviors in order to
lose weight, eat better, exercise more, stop smoking, use less energy, recycle,
etc. [30]. Such experiences have contributed towards the establishment of
concrete design strategies [17].

Applications specific to drivers behavior change are promising, although
in their infancy, and despite evidence of their effectiveness, acceptability, per-
ceived usefulness and possible drawbacks being still unclear (see e.g. [61, 79].
Building on the ubiquity and flexibility of smartphones, a number of appli-
cation have emerged that combine personal mobile sensing and persuasive
strategies, most notably with the aim of promoting a more economic driving
behavior [10, 61, 96, 66, 27]

Other examples of in-car persuasive technologies include conversational
entertainment systems aimed at engaging the driver [93], sensing and pro-
viding feedback on driving mistakes [7], presenting contextual information at
the dashboard to prevent speeding [51] , to mention but a few.

The present vision articulates a novel approach to persuasion that lever-
ages the role of social norms and self-efficacy in order to achieve a change in
behaviour.

2.1. Social norms

Our social life is characterised by norms that manifest themselves as at-
titudinal and behavioural uniformities among people. Social norms can be
viewed as rules and standards that are understood by members of a group,
and that guide and/or constrain their social behaviour [73].

These norms emerge out of interactions with others, they may be either
implicit or explicit, and sanctions for deviating from them come from within
the social network itself [16]. Essentially, social norms are conventions emerg-
ing from a group of people that direct or specify how people must, should

7



or could behave in various situations. Their influence can extend to the
attitudes, beliefs and values held by group members. Social norms can be
seen as a way to maintain stability among the members of a social group or
community [33].

As such, social norms can have a significant influence on individual be-
haviour. In fact, norms are one of the principal ways that social groups
influence individual behaviour and attitudes [73]. How strongly an individ-
ual is affected by the norms of a certain group is influenced by how important
that group is conceived to be for the individual.

2.1.1. Using social network concepts to influence norms

It is widely acknowledged that the use of social networks on desktop or
smartphone devices have shaped our shopping pattern, interaction behaviour,
and education. However, social networks have not comprehensively been
researched or integrated in the driving environment. One reason is that the
enabling technologies are still being developed and/or are still limited (see
sections 1.2.1 and 1.2.1), the other reason is that their road safety potential
has not been fully recognised.

Driving context is fairly different from other contexts as it is very fluid, it
has legal road rules, and requires constant cognitive attention do perform crit-
ical tasks. For example drivers cannot legally use their smartphones whilst
driving. Driving is a critical task and the use of technology could potentially
shift driver’s attention away from the critical task and cause crashes. Driver
distraction, although under reported, is estimated to contribute to 23 percent
of road crashes [74].

On one hand, the use of social networks have been thoroughly investigated
outside the driving context. On the other hand, a large body of research in
ergonomic and human factors have been conducted in the area of Human
Machine Interface (HMI) to minimise driver distraction [74]. However, the
impacts of social network information on driver behaviour has not received
much attention despite its potential to improve mobility, safety and carbon
footprint, a shortcoming that this paper aims to address.

Changing behaviour with the use of real time Human Machine Interface
(HMI) interfaces (e.g. feedback about driving performance) is a psycho-
logically and socially complex problem. Unless the driver already holds a
strong goal (e.g. to be eco-friendly), the feedback will only inform, but will
not necessarily motivate sustainable action. Attitudes, beliefs and values are
learned psychological constructs that have been used successfully to motivate

8



behaviour changes (see e.g. [87]) outside the driving context. Furthermore,
individuals generally respond to the actions and belief of their peers. The
theory of planned behaviour (TPB) is a behavioural science theory which
stipulates that a given behaviour performed by a person depends upon a
combination of particular individual and social factors [4]. In the driving
context, social norms have already had a regulatory role. For example, inten-
tions to drink and drive are influenced by different factors including attitudes,
probability of being caught and social norms [3].

Today, the advent of cooperative systems such V2V combined with ex-
ponential growth of social networks offers unique and unprecedented oppor-
tunities to make a targeted behaviour a contagious social norm transmitted
by V2V communication.

2.2. Self-efficacy

Drivers do not simply react to their immediate environment, but are in-
volved in complex forethought and decision making processes. A substantial
body of converging evidence shows that perceived self-efficacy significantly
influences human self-development, adaptation and change [9]. Self-efficacy
is our belief in our ability to succeed in certain situations. Self-efficacy is a
social cognitive theory in which perceived self-efficacy is a major determinant
of intention. A decision based on misjudgments of our capabilities to reach a
goal could produce detrimental consequences, e.g., proper appraisal of one’s
own efficacy has considerable value. There is no all-purpose measure of per-
ceived self-efficacy [9]. Self-efficacy assessment has not been comprehensively
used in designing feedback in driving situations. The use of self-efficacy pro-
vides valuable information on driver acceptability and driver inclination to
adopt a particular behaviour in all circumstances with the view to assess the
true benefits of the changed behaviour.

2.3. Example: Social sharing to create social norms and improved self-efficacy

The ability of V2V to exchange social information via new forms of in-
vehicle HMI could influence different types of behaviour such as drivers’
speed, eco-driving practice, aggressiveness or other aberrant driving behaviours.
V2V and crowdsourcing mechanisms provide the ability to capture snapshots
of general behaviour of peers (e.g. average speed on school zone). It also pro-
vides the opportunity to identify vehicles violating the social norm. Such new
mechanisms will act as strong incentives for complying and belonging to the
social norm.

9



The HMI, which conveys social information, needs to be carefully crafted
to persuade users to use it. We propose to introduce augmented reality
concepts to convey social cues in cars. The HMI, presented on Head Up
Display (HUD), will augment the driver’s social perception by adding, in
his/her field of view, elements that improve social awareness. We hypothesise
that an augmented reality social display will create a ‘social norm’, which will
persuade, or positively reinforce drivers to adopt behavioural improvement
in more powerful ways, based on the theoretical grounding of the effects of
social norms.

For example, green leaf displays that convey how efficient a driver drives
already exist in cars today. However, they are limited in a way that the
information they convey is not shared amongst peers or other drivers to
create a social norm (c.f. Fig. 2, where the number of green leaves shows
how environmentally responsible other drivers are), nor do they convey how
a driver may have improved over time in comparison to other drivers to
create an improved self-efficacy. Overall, advances in V2V and HMI will
allow us to design more impactful HMIs by leveraging social norms and
self-efficacy theory to improve good driving behaviours, such as eco-driving,
aggressiveness or speeding, from a human perspective.

3. Vision B: Sharing social cues to foster pro-social driving be-
haviours

3.1. The car: an isolating metal shield

Social scientist Leckie refered to cars as a ‘semiprivate metal containers’
[52], because driving a car is not conducive to rich social interactions between
drivers. This is partly due to the car’s metal shield, which tends to isolate
and anonymise drivers, and prevents exchange of social cues. The metal
shield prevents drivers from feeling the social presence of another human
being behind another steering wheel.

Understanding the relationships that drivers form with their car and
places is important to understanding interaction between road users. Cars
and their immediate surrounding have been demarcated as the driver’s own
territory [31]. Territories exist to meet both physical and social needs, while
being temporarily or permanently owned, controlled, marked or personalised,
and potentially defended by occupants or owners [5]. The physical size of
the territory is larger than the size of the vehicle. The overlap of territories

10



can trigger conflicts (e.g. following too closely). The metal shield reinforces
the notion of this territory and could be seen as obstructing social contacts.

Lupton [55, 54] articulates this idea in quite provocative terms: “When
one is driving, one becomes a cyborg, a combination of human and machine.
The notions of individual space, social norms and relationships change to
suit this combination, to the point that drivers tend to humanise cars or,
reversely, to relate to other drivers as machines, thus dehumanising them”.

As a result, at the steering wheel, we all experience an inhibition of our
non-verbal communication capabilities with other road users. We have dif-
ficulty in establishing eye contact, in interpreting facial expressions, and in
using gestures to regulate social interaction; we are limited in our capac-
ity of pointing out our interest, intention or attention to other drivers; we
feel a lack of social and emotional reciprocity. These ‘symptoms’ could well
be the hallmark of autism, and yet represent a common driving situation
that results in drivers often behaving aggressively towards other road users,
misinterpreting others’ intentions.

3.2. The importance of social cues

It is well known that a successful social interaction relies on non-verbal
cues: [6] has estimated that when two people chat, only 30% of the commu-
nication process is verbal, the other 70% is a result of indirect or nonverbal
communication.

Face perception is an important visual cue, which plays a critical role in
social interactions. Perceiving and understanding facial movements changes
due to various type of emotions plays a central role in social communication.

Eye contact is a major social cue to driving safely. Knowing the inten-
tions of other drivers is one of the informal road rules that drivers use to
avoid crashes. Eye contact is a good predictor of attentional focus [6] and
gaze direction plays a crucial role in the initiation and regulation of social
encounters, including the expression of intimacy and dominance [50]. Being
able to make eye contact is arguably one of the major foundations of social
skills.

Gestures, in their many forms and nuances, from gesticulation to formal
sign languages, are tightly related to our capacity of reasoning, remembering
and making sense of the world [60, 59], spatial reasoning, such as when giving
or receiving directions [42], and gesturing while explaining or memorizing
complex procedures is known to lighten the cognitive load in favour of the
competing task [34, 18, 44]. On the other hand, we know very little of how

11



gesturing and non-verbal communication affects the driving task and the
negotiation of shared road resources.

3.3. Social competence

The capacity of effectively and safely negotiating the shared road resource
could be seen as a context specific example of social competence: i.e. the
desirable skills, the status among peers, the ability to form effective relation-
ships and achieve specific goals. Rose-Krasnos proposes a framework that
defines social competence as effectiveness in interaction relative to specific
contexts or domains, both for the self (e.g. personal goals) and the others
(e.g. aspects of competence that involve collaboration and connectedness).

Among the many skills on which social competence is built, e.g. social,
emotional and cognitive abilities [78], an important role is played by social
skills, social communication, interpersonal communication, emotional com-
petence, as well as the ability to respond flexibly to challenges, form effective
relationships and take another’s perspective, as emphasizes e.g. in [86]. All
such skills seem to be compromised or obstructed when driving: for exam-
ple a majority of drivers regard themselves as less risky and more skillful of
the average fellow driver [92]. Despite this, drivers struggle to re-establish
a channel of social communication, and invent means of exchanging non-
verbal cues, using those tools that they have at hand: headlights, hazard
lamps, blinkers, and of course, hand gestures, in what Renge has dubbed
‘roadway interpersonal communication’[75].

3.4. From driver aggression to pro-social Behaviour

Adopting a pro-social behaviour is commonly considered a good driving
practice [72]. Moreover, the inability to express or perceive appropriate social
interactions between drivers could result in aggression, selfish driving and
anti-social behaviour. In other words, driving can make us angry, and anger
at the steering wheel kills.

In a recent survey 50% of drivers admitted to have verbally abused an-
other driver; remarkably 82% of them felt such act to be justified [2]. Driver
aggression can take many forms. Soole and colleagues report that a larger
majority of cases consist of mild aggression (such as verbal abuse, obscene
gestures and tailgating). However, up to 18% of motorists reported severe
aggressions, e.g. having been chased, run off the road, or assaulted [88]. It
is difficult to estimate how aggressive driving contributes to crashes, partly
due to a lack of consistency in the use of terms such as driver aggression,

12



hostile or angry driving, road rage, etc. [24]. Although the cases of road
anger escalating to physical aggression are relatively unusual, several studies
show that aggressive driving is associated with increased risk of crashes [88],
and aggressive behaviour has been reported to be a contributing factor in a
majority of crashes [1].

The reasons why drivers are so prone to getting angry are not limited to
isolation and anonimity. Personal and situational factors, such as age and sex,
previous anger or stress, competitiveness, sensation seeking, anonymity, road
congestion and time pressure, all participate to exacerbate anger episodes
[64].

Subjective interpretations of the behaviour and motivation of others rep-
resent however one major factor. Anger, when articulated, is an important
component of social interaction; it has the function of soliciting cooperation
or apology from a supposedly offending person, or to direct accordingly the
blame of onlookers. In face to face interaction, anger typically cools down as
soon as an acknowledgement is received [68]. In the car, attempts of express-
ing anger and receiving feedback or apology are limited: social cues such as
voice tone or face expressions are unavailable unless overly exaggerated, and
cannot promptly reach other drivers. Similarly, the offending driver will only
receive over-amplified expressions of anger. Such disproportion and the lack
or delay of feedback exacerbates issues of anger [67].

3.5. Sensing and sharing social cues to reduce driver aggression

In our vision, future cars will exploit V2V infrastructure to support and
enhance drivers’ social competence. Existing sensors can detect, interpret
and transmit different forms of emotion [11]. In-vehicle technology combined
with cooperative systems can be used to convey social cues (in this vision,
these cues include eye gaze direction, facial or emotion expression) which can
be collected with in-vehicle sensors.

Sensing technologies for driver behavior and natural user interfaces are
reaching their maturity and are in some cases ready for adoption in industry,
e.g. drowsiness detection and voice activated commands. Other technolo-
gies, though promising as research topics, are still young and immature, but
developing fast. Techniques and tools exist that are capable of classifying
the emotional state of a subject from (combinations of) streams of signals
[19, 13, 12].

Such signals include linguistic and paralinguistic cues in speech [45, 46, 85]
and have been proven to be effective despite of the challenges related to the

13



typically noisy environment in the car [84, 36].
Physiological measures (e.g. heart rate, blood pressure, galvanic skin

response, respiration rate), though generally more invasive, deliver substan-
tially more accurate measures of emotional arousal. MIT’s SmartCar project
compared electromyogram, electrocardiogram, galvanic skin response and
respiration, and trained appropriate recognizers to an impressively accurate
prediction rate[39].

Using facial expressions indicator of emotional arousal is obviously less
accurate, but it has the distinct advantage of being more accessible [83].
Furthermore, it can be combined with other modalities, such as speech, for
better recognizing the emotional state of the driver [41].

Attempts have also been made to recognize context information (e.g.
traffic conditions, presence of pedestrians) and corresponding behavioural
patterns on the drivers’ part (such as eye gaze and body posture) with the
goal of predicting their attentiveness and intentions [95, 94].

Gestures have been explored as a means of controlling the in-vehicle in-
formation system (e.g. [8, 25, 77, 26, 35, 22]). The rationale behind such
explorations is that gesture based interaction could reduce the visual demand
of secondary controls. Note, however, that scientific evidence of an advantage
for gesture based interfaces over more traditional paradigms, with respect to
driver distraction, is still missing [69].

Finally (though without any aim of being comprehensive), eye gaze in
the car has been extensively used in road safety and ITS studies [72]. For
example, previous research has demonstrated that perceiving the eye gaze
of other drivers could influence driver behaviour [71], promoting a pro-social
behavior.

Despite the many challenges and open research topics, natural interfaces,
such as those ones based on gestures and manipulations, could have the po-
tential to foster and reward a pro social behavior. Positive evidence comes
from a domain as far as health care. Tangible User Interfaces [97, 43] have
been found to induce more cooperative and social behaviors in children with
Autism [80] [28], encouraging turn taking and shared attention. Socially
assistive robots have been proposed to foster and facilitate the interaction
between patients and therapists and as a replacement for living pets in pet
therapy [14, 58, 57]. Shape and format of socially assistive robots varies
greatly, from socially proactive robot pets, to ‘intelligent toys’ that reward
the patients when they positively interact with the therapist. El Kaliouby
and colleagues provide a discussion of how affective computing can draw in-

14



sights from the experiences and literature related to social spectrum disorders
such as Autism [47]. Our aim is to achieve analogous results in the car, as
described below.

3.6. Example: technology mediated social cues to restore non-verbal commu-
nication

Fig. 3 illustrates our vision. When approaching a busy urban intersection,
the driver is assisted in the decision making and in the appraisal of the overall
driving situation by devices and displays that enrich the social car.

Ad-hoc sensors, worn by road users or fitted in the vehicle, will recognize a
broad range of context information and socially relevant cues. In this figure,
the pedestrian (A) is attempting to establish eye-contact and is waving a
hand gesture towards the driver. The yellow car (B) is about to turn right,
which can be inferred by patterns of eye-gaze and head pose of the driver.
The red car (D) is conducted by a learner. The cyclist (E) is signalling
her intention to turn left. At the same time, sensors in the car (C, F, H)
are constantly monitoring the behavior and actions of the driver, which will
be transmitted to the other road users and made available via appropriate
interfaces, which can have diverse forms and shapes, ranging from a) haptic
vibrotactile feedback (G) at the steering wheel to b) visual alerts overlayed
to the windscreen (augmented reality), to c) wearable or implanted devices.

The implementation of a scenario such as the one described here involves
the orchestration of cutting edge sensor technology and advanced visual inter-
faces. As we have thoroughly discussed in the previous sections, the building
blocks of this scenario are available, but have rather been used towards the
creation of advanced driver assistance systems, than in the view of supporting
socially oriented in-vehicle information systems.

On the opposite, in pursuing our vision, we rather explore the applica-
tions of such enabling technologies to the collection and sharing of socially
relevant cues. Once the sensors installed in the car have detected the poten-
tial conflicts occurring in the scene, the visual representation depicted in the
figure is created and projected in order to reduce the ambiguity and hence
the risks. Thus, the implementation of the overall vision is the common aim
of several smaller initiatives that we are conducting within our lab, each one
focused at exploring a specific research issue, and each one described in depth
in the the works briefly summarized below.

In [83] we focus on emotion detection as a case study for the recognition
and communication of social cues to enrich the way we communicate and ad-

15



dress conflicts when driving. We argue that although most (if not all) efforts
in the machine learning domain are aimed at achieving higher recognition
rates for those emotions that are more widely and consistently recognized
across cultures, such as anger, disgust, happiness, surprise, fear and sad-
ness. We have argued that in the driving context these emotions make little
sense, and efforts should instead be focused at distinguishing expressions (or
moods) that are desirable or not when driving.

In [89] we further explore the use of surrogate measures such as facial
expression (emotion) and head pose and movements (intention) to infer task
difficulty in a driving situation, showing that a fairly high accuracy can be
obtained even on low cost hardware, such as a personal mobile phone.

Such results help to overcome some of the issues related to the cost of
deployment; it is evident, however, that the development and tuning of sens-
ing technology is challenging, and a lot of work is yet to be done in order to
have affordable and reliable sensors capable of recognizing complex human
behavior.

Even more challenging, however, is the presentation to the driver of such
a massive amount of socially related information. The very decision of what
information is relevant appears not trivial, as we explore in [81, 82] and
discuss further on.

In [71, 72] we explored the possibility of reducing driver aggression by hu-
manising cars in traffic situations by representing other drivers through over-
layed human-like avatars. In [90], we further explore several possible HUD
interfaces in a driving simulator study that aimed at visualizing social cues.
We especially focused on applications aimed at mitigating driver aggression,
each interface was designed to address one or more main contributing fac-
tor of aggressive behavior: anonymity, social isolation, emotional isolation,
competitiveness, territoriality, stress.

The initial results from our experiments provided precious insights into
the technical feasibility of the scenario described here, but also a view of
the complexity of the problem. While a detailed report on such works has
been presented elsewhere [90, 89, 83], it is important to note here that the
successful implementation of our vision for the social car passes through
the understanding of the complex social interactions that occur among road
users before being able to sense, recognize, and ultimately mediate, such
interactions.

16



4. Vision C: Social cars to reduce risky driving

4.1. Significant over-representation of young males in road crashes

Worldwide, over a million people are killed and an additional 50 million
are seriously injured on roads annually [76]. The heaviest toll is paid by
young drivers [70]. Young male drivers are at a substantially higher risk of
committing and repeating speeding offences [99] and being involved in speed-
related fatal accidents [53]. Young male drivers are also more likely to be
distracted while driving, especially through mobile phone use [15] an issue
authorities struggle to address.

4.2. Boredom prone young male drivers are driven to distraction and risk

There are many underlying factors for the over-representation of young
males in motor vehicle crashes. One of those factors is that young males
often score highly on sensation seeking measures, and therefore tend to be
prone to boredom ([100][98][40]).

Boredom directly correlates to sensation seeking, but has only received
limited attention to date [38]. Boredom is defined as a ‘state of relatively
low arousal and dissatisfaction, which is attributed towards an inadequately
stimulating environment’ (p.3 [63]). In the driving context, boredom leads
to the following problem: If the driving environment is not providing enough
stimulation, young male drivers tend to seek sensations by taking risks (see
Fig. 4); this includes increasing their speed or diverting their attention away
from the driving task [32], e.g. by texting.

The challenge in addressing this core problem is the fact that proneness
to boredom (and sensation seeking behaviour) is a hardwired personality
factor in young males [38]. This means that it cannot be changed and that
existing road safety strategies such as education programs, punitive fines or
awareness campaigns are conceptually flawed in addressing it. Indeed, the
paucity of research on strategies that tackle driver boredom left authorities
unable to deal with technologies and devices that enter the safety critical car
space.

4.3. From foe to friend: Using social computing devices to aid in safer driving
behaviour

Authorities and the road safety community often perceive new social tech-
nologies that enter our cars first and foremost as a threat that aggravates
risky driving behaviours by distracting drivers. Alarmingly, many road safety

17



researchers tend to focus on their safety impacts only after they have started
to be used inside cars, cf. smartphones and texting [15]. Regardless of puni-
tive strategies, new technologies will continue to be used within the car, es-
pecially by young males who are early adopters. New distracting ubiquitous
computing devices fulfilling our insatiable need for social connectedness will
enter the car space in the foreseeable future, which could pose an even greater
challenge to road safety, e.g., new wearable computing devices, Augmented
Reality (AR) glasses (e.g. Google Project Glass) and Head Up Displays
(e.g. Pioneer CyberNavi). To address this imminent road safety problem, a
conceptual breakthrough is needed. Instead of indiscriminately and futilely
rejecting and demonising such technologies, this vision aims to encourage the
discovery of new ways to capitalise their usage towards bringing about safer
driving practices.

4.4. Curing boredom by safely providing stimuli when needed

Above, it was established that: 1. Young male drivers are prone to bore-
dom; 2. Their need for stimulation is hardwired and not modifiable; 3.
Boredom leads to risky driving behaviour; and 4. Existing strategies are
conceptually flawed and ineffective. Consequently, there is an urgent need to
find innovative ways to provide alternative stimuli while driving when needed
and when it is safe to do so with the view to curb the road toll. At the core
of this visionary research lie the following questions: How can innovative
technologies be designed to make safer driving behaviours equally or more
pleasurable and stimulating than risky driving behaviours? Can the alter-
native stimuli be grounded in our need for social connectedness? Can they
be designed in a way that they replace seeking risky driving stimuli, hence
reducing risky driving (Fig. 5)?

This concept is innovative in that it represents a paradigm shift. It is
neither oblivious to road safety risks nor is it using a patronising approach
by telling young male drivers what not to do without offering them an alter-
native to satisfy their need. It follows the hypothesis, that providing safe,
driving-related and emphpleasurable stimuli through digital, social technolo-
gies in the car can replace the urge for risky driving behaviours in young
male drivers. The theoretical grounding is based on the premise that these
additional stimulations can break the boredom, hence diverting the atten-
tion back towards the safe driving task and away from seeking stimulations
through risk taking. Overall, this as yet unexplored concept articulates the
notion of pleasure and safety in one technology intervention with the view

18



to reduce injuries and fatalities. This notion was developed through a pilot
study [82] and evaluated in the road safety domain [81].

Conceptually, this approach is similar to the Australian Government’s
“Swap It! Don’t stop it!” campaign (www.swapit.gov.au), which encourages
Australians to swap bad habits with healthier alternatives, or the Volkswagen
Fun Theory initiative (www.thefuntheory.com), which presents ideas that
make good behaviour more pleasurable or appealing than bad behaviour.
For example, the Speed Camera Lottery rewards drivers complying with the
speed limit with the chance to win a lottery financed by fines paid by speed
offenders. By making good behaviour fun, it reduced average speeds from 32
km/h to 25 km/h.

The hypothesis presented above leads to the following research ques-
tions: What are driving-related and pleasurable stimuli for different driver
archetypes, particularly boredom prone young males? How can new tech-
nologies be designed to safely provide pleasurable stimuli at the right time?
Do pleasurable stimuli have safety benefits by replacing the urge for risky
driving behaviours?

4.5. The ‘what’ brings pleasure and safety together

The concepts of pleasure and safety have conventionally been portrayed
to pull apart from each other (Fig. 6). Car manufacturers and the research
community around Automotive User Interfaces aim to unite the pleasure of
driving and road safety by focusing on the question as to how a driver can
safely interact (output/input) with various types of data or information with-
out causing driver distraction (Fig. 6). The question how to safely output or
input information undoubtedly pushes the technological advancements, mak-
ing in-car Human-Machine (HMI) and Human-Computer interaction (HCI)
safer. However, the ‘How’ is only part of the solution. The actual infor-
mation or applications (the ‘What’ in Fig. 6) that form the basis of in-car
HMI/HCI research have not changed much in recent years. They generally
include tasks such as writing/reading SMS, emails, or more recently tweets
and social media status updates; dialling or making phone calls, selecting
from lists of the in-car entertainment system; operating the navigation sys-
tem; or, exploring points of interest. Little attention has been paid to new
types of content and applications.

19



4.6. Example: Rewarding achievements

Online platforms often use socially grounded gamification or ratings to
increase fun and engagement, motivate users to come back to use the sys-
tem, or enforce certain behaviours or etiquettes. E.g., in Foursquare, users
can claim mayorships, unlock badges, receive special offers & rewards from
retailers while also tracking against friends via a leader board [20]; on eBay,
buyers and sellers rate each other to encourage fair trading.

This vision seeks to translate these concepts (without being limited to
gamification and rating systems) to road safety in order to stimulate drivers
with engaging, driving related tasks when they are being understimulated
and to, e.g., motivate drivers to drive more safely and courteously.

A simple example to illustrate this concept is depicted in Fig. 7. It shows
an augmented reality application that allows drivers to a) rate each other,
b) accomplish save driving related achievements by being rated and c) view
each other’s achievements. Adding such a layer of playful engagement obvi-
ously requires save interfaces for inputting and outputting such information,
but conceptually it could stimulate the driver with a playful, driving-related
task that keeps them ’distracted’ (in a good way) from other risky driving
decisions, such as speeding.

As mentioned, this presents just one example to illustrate the concept.
However, we will explore this innovative approach to road safety in more
detail in the future. As demonstrated, the literature in social and psychology
research provides the basis to pursue this vision in greater detail.

5. Conclusion

The social cost of road injury and fatalities is still unacceptable. The
driver is often the main responsible for road crashes, therefore changing the
driver behaviour is one of the most important and most challenging priority
in road transport.

Human’s craving for social connectedness is alleviated with the increasing
use social networks. However we have shown that social connectedness is
somehow restricted in a car. Social theories stipulate that one of the most
sustainable ways to change individual behaviour is to influence peers. Social
norms are used to provide a “scheme” on the appropriateness of our own
behaviour. There is a plethora of theoretical evidences demonstrating that
social concepts could influence driver behaviour. Social networks, used and

20



adapted in the driving context, can harbor a flow of desirable moral values
such as safety or green driving behaviour.

Tackling crucial road transport problems such as agressivity, eco-driving
and safety with social information, conveyed with V2X is an innovative ap-
proach. We introduced a simple but powerful concept. It consists of revealing
“good” aspects of the driver behaviour to other drivers and make it conta-
gious. The use of self-efficacy , social norms and gamification theories could
increase the likelihood of a widespread adoption of such “good” behaviour.

The use of such social information could make driving more humane.
Hence, we have shown how V2V could convey and establish social norms with
the view to positively influence driver’s behaviour. Specifically we showed
how V2V could (i) break the vehicle’s ‘social’ shield, (ii) provide a better
perception of ‘presence’ of other road users, (iii) shift our perception of
acceptable driving practices (iv) reduce agressivity and (v) improve road
safety.

Cars and V2V have never been on the spotlight for being able to change
behaviour and spread such moral values. The idea that one’s behavior and
actions can influence not only the drivers in the vicinity but also the others
connected in our social network is a very powerful concept worth investigating
for future research. Testing and validating such idea needs to be conducted
in naturalistic settings as real-time in-vehicle information is likely to impact
positively or negatively on driver behaviour. The net benefit of our proposed
concepts is subjected to driver acceptance and market penetration.

In the future, our behavior, actions, beliefs and habits are likely to be
largely more influenced and impacted by social networks and social media
supported by cooperative systems than we ever could have imagined.

References

[1] AAAFoundation.org, Aggressive Driving: Research Up-
date, Technical Report, AAAFoundation.org, 2009. URL:
https://www.aaafoundation.org/sites/default/files/

AggressiveDrivingResearchUpdate2009.pdf.

[2] AAMI, Annual Road Safety Index, 2011. URL:
http://www.aami.com.au/sites/default/files/fm/news/AAMI

Crash Index 2011 FINAL.pdf.

21



[3] L. Äberg, Drinking and driving: Intentions, attitudes, and social norms
of Swedish male drivers, Accident Analysis and Prevention 25 (1993)
289–296. doi:http://dx.doi.org/10.1016/0001-4575(93)90023-P.

[4] I. Ajzen, From intentions to actions: A theory of planned behaviour,
From intentions to actions: A theory of planned behaviour, Springer
Verlag, 1985, pp. 11–40.

[5] I. Altman, M. Chemers, Culture and environment, Brooks/Cole, Mon-
terey, CA, 1980.

[6] M. Argyle, Social Interaction. 1969., London: Methuen, 1969.

[7] E. Arroyo, S. Sullivan, T. Selker, CarCoach: a polite and effec-
tive driving coach, in: Proceedings of ACM CHI 2006 Conference
on Human Factors in Computing Systems, volume 2, pp. 357–362.
doi:10.1145/1125451.1125529.

[8] K.M. Bach, M.M.G. Jæ ger, M.B.M. Skov, N.G.N. Thomassen, You
can touch, but you can’t look: interacting with in-vehicle systems, in:
Proceedings of the twenty-sixth annual SIGCHI conference on Human
factors in computing systems, CHI ’08, ACM, New York, NY, USA,
2008, pp. 1139–1148. doi:10.1145/1357054.1357233.

[9] A. Bandura, Self-Efficacy, John Wiley and Sons, Inc., 2010.
doi:10.1002/9780470479216.corpsy0836.

[10] A. Bergmans, S. Shahid, DriveRS: An In-Car Persuasive System for
Making Driving Safe and Fun, in: A. Nijholt, T. Romão, D. Reidsma
(Eds.), Advances in Computer Entertainment, volume 7624, Springer
Berlin/Heidelberg, 2012, pp. 469–472. doi:10.1007/978-3-642-34292-
9 37.

[11] K. Boehner, R.R. Depaula, P. Dourish, P. Sengers, A.C. Zaidan, U.C.
Irvine, How emotion is made and measured, Int. J. Hum.-Comput.
Stud. 65 (2007) 275–291. doi:10.1016/j.ijhcs.2006.11.016.

[12] S. Brave, C. Nass, K. Hutchinson, Computers that care: investigating
the effects of orientation of emotion exhibited by an embodied computer
agent, International Journal of Human-Computer Studies 62 (2005)
161–178. doi:10.1016/j.ijhcs.2004.11.002.

22



[13] S. Brave, C. Nass, J. Preece, D. Maloney-Krichmar, The human-
computer interaction handbook, L. Erlbaum Associates Inc., Hillsdale,
NJ, USA, 2003, pp. 596–620.

[14] J. Broekens, M. Heerink, H. Rosendal, Assistive social robots in elderly
care: a review, Gerontechnology (2009).

[15] CARRS-Q, Distraction and Inattention Fact Sheet - Mo-
bile Pone Use & Distraction While Driving, Techni-
cal Report, CARRS-Q, Brisbane, QLD, 2012. URL:
http://www.carrsq.qut.edu.au/publications/corporate/

mobile phones and distraction fs.pdf.

[16] R.B. Cialdini, M.R. Trost, Social influence: social norms, conformity
and compliance, in: D.T. Gilbert, S.T. Fiske, G. Lindsey (Eds.), The
Handbook of social psychology, volume II, McGraw-Hill, 1998, pp. 151–
192.

[17] S. Consolvo, D.W. McDonald, J.A. Landay, Theory-driven design
strategies for technologies that support behavior change in everyday
life, in: Proceedings of the SIGCHI Conference on Human Factors
in Computing Systems, ACM, Boston, MA, USA, 2009, pp. 405–414.
doi:10.1145/1518701.1518766.

[18] S.W. Cook, Z. Mitchell, S. Goldin-Meadow, Gestur-
ing makes learning last, Cognition 106 (2008) 1047–1058.
doi:10.1016/j.cognition.2007.04.010.

[19] R. Cowie, E. Douglas-Cowie, N. Tsapatsoulis, G. Votsis, S. Kollias,
W. Fellenz, J.G. Taylor, Emotion recognition in human-computer
interaction, Signal Processing Magazine, IEEE 18 (2001) 32–80.
doi:10.1109/79.911197.

[20] H. Cramer, M. Rost, L.E. Holmquist, Performing a check-in: emerging
practices, norms and ’conflicts’ in location-sharing using foursquare,
in: Proceedings of the 13th International Conference on Human
Computer Interaction with Mobile Devices and Services - Mobile-
HCI ’11, ACM Press, New York, New York, USA, 2011, p. 57.
doi:10.1145/2037373.2037384.

23



[21] S. Demmel, A. Lambert, D. Gruyer, A. Rakotonirainy, E. Monacelli,
Empirical IEEE 802.11p performance evaluation on test tracks, in:
IEEE Intelligent Vehicles Symposium 2012, IEEE, Alcala de Henares,
Spain, 2012, pp. 837–842.

[22] T. Döring, D. Kern, P. Marshall, M. Pfeiffer, J. Schöning, V. Gruhn,
A. Schmidt, Gestural interaction on the steering wheel: reducing the
visual demand, in: Proceedings of the 2011 annual conference on Hu-
man factors in computing systems, CHI ’11, ACM, New York, NY,
USA, 2011, pp. 483–492. doi:10.1145/1978942.1979010.

[23] P. Dourish, Where the Action Is: The Foundations of Embodied Inter-
action, new editio ed., The MIT Press, 2004.

[24] C.S. Dula, E. Geller, Risky, aggressive, or emotional driving: Address-
ing the need for consistent communication in research, Journal of Safety
Research 34 (2003) 559–566. doi:10.1016/j.jsr.2003.03.004.

[25] R. Ecker, V. Broy, A. Butz, A. De Luca, pieTouch: a direct
touch gesture interface for interacting with in-vehicle information
systems, in: Proceedings of the 11th International Conference on
Human-Computer Interaction with Mobile Devices and Services, Mo-
bileHCI ’09, ACM, New York, NY, USA, 2009, pp. 22:1—-22:10.
doi:10.1145/1613858.1613887.

[26] R. Ecker, V. Broy, K. Hertzschuch, A. Butz, Visual cues sup-
porting direct touch gesture interaction with in-vehicle information
systems, in: Proceedings of the 2nd International Conference on
Automotive User Interfaces and Interactive Vehicular Applications,
AutomotiveUI ’10, ACM, New York, NY, USA, 2010, pp. 80–87.
doi:10.1145/1969773.1969788.

[27] R. Ecker, P. Holzer, V. Broy, A. Butz, EcoChallenge: a race for effi-
ciency, Proceedings of the 13th International . . . (2011).

[28] W. Farr, N. Yuill, H. Raffle, Social benefits of a tangible user in-
terface for children with Autistic Spectrum Conditions., Autism :
the international journal of research and practice 14 (2010) 237–52.
doi:10.1177/1362361310363280.

24



[29] J. Fleiter, A. Lennon, B. Watson, How do other people influence your
driving speed? Exploring the ’who’and the ’how’of social influences
on speeding from a qualitative perspective, . . . research part F: traffic
psychology and . . . (2010).

[30] B.J. Fogg, Persuasive technology: using computers to change what we
think and do, Morgan Kaufmann, San Francisco, CA, 2002.

[31] G. Fraine, S.G. Smith, L. Zinkiewicz, R.L. Chapman, M.C. Sheehan,
At home on the road? Can drivers’ relationships with their cars be
associated with territoriality?, Journal of Environmental Psychology
27 (2007) 204–214.

[32] R. Fuller, Towards a general theory of driver behaviour.,
Accident; analysis and prevention 37 (2005) 461–472.
doi:10.1016/j.aap.2004.11.003.

[33] A. Gangemi, Norms and plans as unification criteria for social collec-
tives, Autonomous Agent Mutli-Agent System 17 (2008) 70–112.

[34] S. Goldin-Meadow, Hearing Gesture: How Our Hands Help Us Think,
Belknap Press of Harvard University Press, 2005.

[35] I.E. González, J.O. Wobbrock, D.H. Chau, A. Faulring, B.A. Myers,
Eyes on the road, hands on the wheel: thumb-based interaction tech-
niques for input on steering wheels, in: Proceedings of Graphics In-
terface 2007, GI ’07, ACM, New York, NY, USA, 2007, pp. 95–102.
doi:10.1145/1268517.1268535.

[36] M. Grimm, K. Kroschel, H. Harris, C. Nass, B. Schuller, G. Rigoll,
T. Moosmayr, On the Necessity and Feasibility of Detecting a Driver’s
Emotional State While Driving, in: A. Paiva, R. Prada, R. Picard
(Eds.), Affective Computing and Intelligent Interaction, volume 4738
of Lecture Notes in Computer Science, Springer Berlin / Heidelberg,
2007, pp. 126–138.

[37] D. Gruyer, S. Demmel, B. D’Andrea-Novel, A. Lambert, A. Rako-
tonirainy, Simulation architecture for the design of Cooperative Col-
lision Warning systems, in: 2012 15th International IEEE Confer-
ence on Intelligent Transportation Systems, IEEE, 2012, pp. 697–703.
doi:10.1109/ITSC.2012.6338733.

25



[38] J. Harvey, S. Heslop, N. Thorpe, The categorisation of drivers in rela-
tion to boredom, Transportation Planning and Technology 34 (2011)
51–69. doi:10.1080/03081060.2011.530829.

[39] J. Healey, R. Picard, SmartCar: detecting driver stress, in: Pattern
Recognition, 2000. Proceedings. 15th International Conference on, vol-
ume 4, pp. 218 –221 vol.4. doi:10.1109/ICPR.2000.902898.

[40] A.B. Hill, Work variety and individual differences in occupa-
tional boredom, Journal of Applied Psychology 60 (1975) 128–131.
doi:10.1037/h0076346.

[41] S. Hoch, F. Althoff, G. McGlaun, G. Rigoll, Bimodal fusion of emo-
tional data in an automotive environment, in: Acoustics, Speech,
and Signal Processing, 2005. Proceedings. (ICASSP ’05). IEEE In-
ternational Conference on, volume 2, pp. ii/1085 – ii/1088 Vol. 2.
doi:10.1109/ICASSP.2005.1415597.

[42] A. Hostetter, M. Alibali, Visible embodiment: Gestures as simulated
action, Psychonomic Bulletin and Review 15 (2008) 495–514.

[43] H. Ishii, B. Ullmer, Tangible bits: towards seamless inter-
faces between people, bits and atoms, in: CHI ’97: Proceed-
ings of the SIGCHI conference on Human factors in comput-
ing systems, ACM, New York, NY, USA, 1997, pp. 234–241.
doi:http://doi.acm.org/10.1145/258549.258715.

[44] J.M. Iverson, S. Goldin-Meadow, Why people gesture when they
speak., Nature 396 (1998).

[45] C. Jones, I.M. Jonsson, Using Paralinguistic Cues in Speech to Recog-
nise Emotions in Older Car Drivers, in: C. Peter, R. Beale (Eds.),
Affect and Emotion in Human-Computer Interaction, volume 4868 of
Lecture Notes in Computer Science, Springer Berlin / Heidelberg, 2008,
pp. 229–240.

[46] C.M. Jones, I.M. Jonsson, Performance analysis of acoustic emotion
recognition for in-car conversational interfaces, in: Proceedings of the
4th international conference on Universal access in human-computer
interaction: ambient interaction, UAHCI’07, Springer-Verlag, Berlin,
Heidelberg, 2007, pp. 411–420.

26



[47] R. el Kaliouby, R. Picard, S. Baron-Cohen, Affective computing and
autism., Annals of the New York Academy of Sciences 1093 (2006)
228–48. doi:10.1196/annals.1382.016.

[48] A. Kerckhoffs, La cryptographie militaire, Journal des sciences mili-
taires IX (1883) 5–38.

[49] D. Kirsh, P. Maglio, On distinguishing epistemic from pragmatic
action, Cognitive Science 18 (1994) 513–549. doi:10.1016/0364-
0213(94)90007-8.

[50] C.L. Kleinke, Gaze and eye contact: a research review., Psychological
bulletin 100 (1986) 78–100.

[51] M. Kumar, T. Kim, Dynamic speedometer: dashboard redesign to
discourage drivers from speeding, in: Proceedings of ACM CHI 2005
Conference on Human Factors in Computing Systems, volume 2, pp.
1573–1576. doi:http://doi.acm.org/10.1145/1056808.1056969.

[52] G.J. Leckie, J. Hopkins, The public place of central libraries: Findings
from Toronto and Vancouver, The Library Quarterly 72 (2002) 326–
372.

[53] C. Liu, C.L. Chen, R. Subramanian, D. Utter, Analysis of
Speeding-Related Fatal Motor Vehicle Traffic Crashes, Technical
Report, National Center for Statistics and Analysis, 2005. URL:
http://trid.trb.org/view.aspx?id=763754.

[54] D. Lupton, Monsters in Metal Cocoons: ‘Road Rage’
and Cyborg Bodies, Body & Society 5 (1999) 57–72.
doi:10.1177/1357034X99005001005.

[55] D. Lupton, Road rage: drivers’ understandings and experiences, Jour-
nal of Sociology 38 (2002) 275–290. doi:10.1177/144078302128756660.

[56] P.P. Maglio, D. Kirsh, Epistemic Action Increases With Skill, in: In
Proceedings of the Eighteenth Annual Conference of the Cognitive Sci-
ence Society, Erlbaum, 1996, pp. 391–396.

[57] P. Marti, M. Bacigalupo, C. Mennecozzi, L. Giusti, T. Shibata,
Socially Assistive Robotics in the Treatment of Behavioural and

27



Psychological Symptoms of Dementia, in: The First IEEE/RAS-
EMBS International Conference on Biomedical Robotics and
Biomechatronics, 2006. BioRob 2006., IEEE, 2006, pp. 483–488.
doi:10.1109/BIOROB.2006.1639135.

[58] P. Marti, L. Giusti, M. Bacigalupo, Dialogues beyond words, Interac-
tion Studies (2008).

[59] D. Mcneill, Hand and Mind: What Gestures Reveal about Thought,
University Of Chicago Press, 1992.

[60] D. McNeill, Gesture and thought., University of Chicago Press, 2005.

[61] A. Meschtscherjakov, D. Wilfinger, T. Scherndl, M. Tscheligi, Accep-
tance of future persuasive in-car interfaces towards a more economic
driving behaviour, in: Proceedings of the 1st International Conference
on Automotive User Interfaces and Interactive Vehicular Applications
- AutomotiveUI ’09, ACM Press, New York, New York, USA, 2009,
p. 81. doi:10.1145/1620509.1620526.

[62] J.A. Michon, A critical view of driver behavior models: What do we
know, what should we do?, in: L. Evans, R.C. Schwing (Eds.), Human
behavior and traffic safety, Plenum Press, 1985, pp. 485–520.

[63] W.W.L. Mikulas, S.J.S. Vodanovich, The essence of boredom., The
Psychological Record 43 (1993) 3–12.

[64] S.R. O’Brien, R.S. Tay, B.C. Watson, Situational factors contributing
to the expression of aggression on the roads, IATSS Research 28 (2004)
101–107.

[65] OECD, Road safety: Impact of new technologies, OECD Publications,
Paris, France, 2003.

[66] T. Pace, S. Ramalingam, D. Roedl, Celerometer and Idling Reminder
: Persuasive Technology for School Bus, in: Extended Abstracts CHI
2007, pp. 2085–2090. doi:10.1145/1240866.1240954.

[67] B. Parkinson, Anger on and off the road, British Journal of Psychology
92 (2001) 507–526. doi:10.1348/000712601162310.

28



[68] B. Parkinson, Emotions in direct and remote social interaction: Get-
ting through the spaces between us, Computers in Human Behavior 24
(2008) 1510–1529. doi:10.1016/j.chb.2007.05.006.

[69] C. Pickering, A review of automotive human machine interface tech-
nologies and techniques to reduce driver distraction, . . . and Technology
. . . (2007).

[70] Queensland Department of Transport and Main
Roads, Speeding information, 2012. URL:
http://www.tmr.qld.gov.au/Safety/Driver-guide/Speeding.aspx.

[71] A. Rakotonirainy, F. Feller, N. Haworth, In-Vehicle Avatars to Elicit
Social Response and Change Driving Behaviour, International Journal
of Technology and Human Interaction (IJTHI) 5 (2009) 80–104.

[72] A. Rakotonirainy, F. Feller, N.L. Haworth, Using in-vehicle avatars to
prevent road violence, in: Pervasive 2008, Sydney, Australia, pp. 70–74.

[73] A. Rakotonirainy, P. Obst, S.W. Loke, Socially aware computing con-
structs, IJSHC 1 (2012) 375–395.

[74] M.A. Regan, J.D. Lee, K. Young (Eds.), Driver distraction: Theory,
Effects and Mitigation, CRC Press, 2009.

[75] K. Renge, Effect of driving experience on drivers’ decoding process of
roadway interpersonal communication., Ergonomics 43 (2000) 27–39.
doi:10.1080/001401300184648.

[76] E.D. Richter, T. Berman, L. Friedman, G. Ben-David, Speed, road
injury, and public health., Annual review of public health 27 (2006)
125–52. doi:10.1146/annurev.publhealth.27.021405.102225.

[77] H. Richter, R. Ecker, C. Deisler, A. Butz, HapTouch and the 2+1
state model: potentials of haptic feedback on touch based in-vehicle
information systems, in: Proceedings of the 2nd International Confer-
ence on Automotive User Interfaces and Interactive Vehicular Applica-
tions, AutomotiveUI ’10, ACM, New York, NY, USA, 2010, pp. 72–79.
doi:10.1145/1969773.1969787.

29



[78] L. Rose-Krasnor, The Nature of Social Competence: A Theoreti-
cal Review, Social Development 6 (1997) 111–135. doi:10.1111/j.1467-
9507.1997.tb00097.x.

[79] H. Rouzikhah, M. King, A. Rakotonirainy, Examining the effects of
an eco-driving message on driver distraction, Accident Analysis and
Prevention 50 (2013) 975–983. doi:10.1016/j.aap.2012.07.024.

[80] B. Scassellati, How social robots will help us to diagnose, treat, and
understand autism, Robotics research (2007).

[81] R. Schroeter, A. Rakotonirainy, The future shape of digital cars, in:
2012 Australasian Road Safety Research, Policing and Education Con-
ference, Wellington Convention Centre, Wellington, NZ, pp. 4–6.

[82] R. Schroeter, A. Rakotonirainy, M. Foth, The social car : new in-
teractive vehicular applications derived from social media and urban
informatics, in: 4th International Conference on Automotive User In-
terfaces and Interactive Vehicular Applications (AutomotiveUI 2012),
ACM, Portsmouth, New Hampshire, 2012, pp. 107–110.

[83] R. Schroeter, A. Soro, A. Rakotonirainy, Social Cars : Sensing , Gath-
ering , Sharing and Conveying Social Cues to Road Users, in: B. Guo,
D. Riboni, H. Peizhao (Eds.), Creating Personal, Social and Urban
Awareness through Pervasive Computing, IGI Global, 2013.

[84] B. Schuller, M. Lang, G. Rigoll, Schuller B., Recognition of Spon-
taneous Emotions by Speech within Automotive Environment, in:
Tagungsband Fortschritte der Akustik - DAGA 2006, pp. 57–58.

[85] B. Schuller, S. Reiter, R. Muller, M. Al-Hames, M. Lang, G. Rigoll,
Speaker Independent Speech Emotion Recognition by Ensemble Clas-
sification, in: Multimedia and Expo, 2005. ICME 2005. IEEE Interna-
tional Conference on, pp. 864–867. doi:10.1109/ICME.2005.1521560.

[86] M. Semrud-Clikeman, Social competence in children, Springer, 2007.

[87] M. Shipworth, Motivating Home Energy Action, Australian Green-
house Office (2000).

30



[88] D. Soole, A. Lennon, B. Watson, C. Bingham, Towards a comprehen-
sive model of driver aggression: a review of the literature and directions
for the future, in: C.N. Ferraro (Ed.), Traffic Safety, Nova Science Pub-
lishers, Inc., 2011, pp. 69—-97.

[89] A. Soro, A. Rakotonirainy, Automatic Inference of Driving Task De-
mand from Visual Cues of Emotion and Attention, in: 2nd Interna-
tional Conference on Human Factors in Transportation.

[90] A. Soro, S. Wollstädter, A. Rakotonirainy, Advanced In-Vehicle Appli-
cations to Mitigate Driver Aggression, in: 2nd International Conference
on Human Factors in Transportation.

[91] S. Stradling, Car driver speed choice in Scotland, Ergonomics (2007).

[92] O. Svenson, Are we all less risky and more skillful than our fel-
low drivers?, Acta Psychologica 47 (1981) 143–148. doi:10.1016/0001-
6918(81)90005-6.

[93] J. Tester, B.J. Fogg, M. Maile, CommuterNews: a prototype of
persuasive in-car entertainment, in: CHI 00 CHI 00 extended
abstracts on Human factors in computing systems, pp. 24–25.
doi:10.1145/633292.633309.

[94] M.M. Trivedi, S.Y. Cheng, Holistic Sensing and Active Displays
for Intelligent Driver Support Systems, Computer 40 (2007) 60–68.
doi:10.1109/MC.2007.170.

[95] M.M. Trivedi, T. Gandhi, J. McCall, Looking-In and Looking-Out
of a Vehicle: Computer-Vision-Based Enhanced Vehicle Safety, IEEE
Transactions on Intelligent Transportation Systems 8 (2007) 108–120.
doi:10.1109/TITS.2006.889442.

[96] J. Tulusan, T. Staake, E. Fleisch, Providing eco-driving feedback
to corporate car drivers: what impact does a smartphone appli-
cation have on their fuel efficiency?, in: Conference on Ubiqui-
tous Computing, ACM, Pittsburgh, Pennsylvania, 2012, pp. 212–215.
doi:10.1145/2370216.2370250.

31



[97] B. Ullmer, H. Ishii, Emerging frameworks for tangi-
ble user interfaces, IBM Syst. J. 39 (2000) 915–931.
doi:http://dx.doi.org/10.1147/sj.393.0915.

[98] S.J. Vodanovich, J.C. Wallace, S.J. Kass, A confirmatory approach
to the factor structure of the Boredom Proneness Scale: Evidence for
a two-factor short form, Journal of Personality Assessment 85 (2005)
295–303. doi:10.1207/s15327752jpa8503 05.

[99] B.C. Watson, A. Watson, V. Siskind, J.J. Fleiter, Characteristics and
predictors of high-range speeding offenders, in: 2009 Australasian Road
Safety Research, Policing and Education Conference : Smarter, Safer
Directions, Engineers Australia, Sydney Convention and Exhibition
Centre, Sydney, New South Wales, 2009.

[100] M. Zuckerman, S. Eysenck, H.J. Eysenck, Sensation seeking in England
and America: cross-cultural, age, and sex comparisons., Journal of con-
sulting and clinical psychology 46 (1978) 139–149. doi:10.1037/0022-
006X.46.1.139.

32



List of Figures

Figure 1: The social awareness layers around the social car

33



Figure 2: The green leaf display showing consumption of other drivers

Figure 3: Sharing social cues to reduce driver aggression

34



Figure 4: Under and Over stimulation - boredom and risk

Figure 5: Under and Over stimulation - Intervention

35



Figure 6: The what that brings pleasure and safety together

Figure 7: Rating other drivers and receiving badges for good driver behaviour

36