Exponential technological growth

”We are energized by the great power of technological impact on us.
We are intimidated by the magnitude of problems it creates or alerts
us to.”

Herbert Simon, Nobel Prize Winner 1978, Economics (Boutellier and
Heinzen, 2014, p.1)

In 1965 Gordon Moore, the co-founder of Intel, predicted that in the future
processing power would double every 18 months. ‘Moore’s Law’ is still
referred to today to predict the growth of technology (Loveridge, 1990).
Revolutionary technological developments have a large impact on every
aspect of human life. Technologies such as 3G and 4G, smartphones and the
World Wide Web provide numerous benefits and have simplified our lives.
We now have easy access to information and education, financial
transactions can be completed online, and world trade business is conducted
faster. But the exponential growth of technology is also causing difficulties
(Loveridge, 1990). The McKinsey Global Institute’s report ‘Disruptive
technologies’ of 2013 explores how technological changes transform the way
we live and work, stating:

“Business leaders and policy makers need to identify potentially
disruptive technologies, and carefully consider their potential,
before these technologies begin to exert their disruptive powers in
the economy and society.” (Manyika et al., 2013, preface)

The report analyses technological developments in automation, the internet
of things, cloud technology, advanced robotics, energy storage, 3D printing
and more. Predicted changes in established norms and technological
developments will force society to adapt (Manyika et al., 2013). One discipline
highly influenced by these technological developments is the automotive

16

industry. A selection of the most popular current and future technological
developments in the automotive industry is presented in Table 1 below.

Development

Connected cars,

Vehicle to vehicle (V2V) communication

Real-time internet of things (IoT) connectivity

Increasing number of sensors implemented

Gesture control systems

Voice recognition

Electric cars

In-car connectivity:

Internet connection

Integration of social media

Integration of mobile devices

Augmented reality included into the interface

Partially autonomous driving

Driver assistance features

Fully autonomous driving

Current/Future Source

Sophistication of in-car technologies Current Manyika et al., 2013

Disruptive Technologies

Current and
future

Current

Current

Current
future

Current

Current

Current

Future

Intland, 2016; Kalogianni,
2016; Karlsson, Ahn and
Choi, 2017

Manyika et al., 2013

Ropert, 2014

Manyika et al., 2013; The
Guardian view of the car
industry: an electric future,
2017

Brauer, 2015;
Kalogianni, 2016;
Karlsson, Ahn and Choi, 2017

Ropert 2014

Kalogianni, 2016

Karlsson, Ahn and Choi, 2017

• Car sharing Current and Harris, 2017;

and

• Shared mobility future

Manyika et al., 2013

Table 1 Most popular current and future technological developments in the automotive
industry

1.1 Developments in the Automotive Industry

The main focus of automotive research on basic functional requirements from
1886-1919, changed from 1920-1930 with the dawn of human factors.
Increasing speed, growing numbers of vehicles on roads and increasing
travel times called for a shift in focus towards safety from 1940-1949. Further
developments in the 1960s included 3-point-seatbelts and research on
mental workload and driver fatigue. In the 1970s and 80s emphasis on driver

17

comfort grew. With an increasing number of functions in the automobile and
concentration on the user experience the focus of research shifted towards
the driving experience. The development of multimedia interfaces and
navigation systems was followed by lane-keeping, breaking and blind-spot-
warning systems in the 2000s, a significant step towards further automation
of automobiles (Akamatsu, Green and Bengler, 2013; Miller, 2001; Normark
and Gkouskos, 2012). Today, the connected car has improved
communication, connection and networking (Manyika et al., 2013), major
steps towards autonomous driving (Figure 1).

Figure 1 Technological developments in the present and future (Karlsson, Ahn and Choi,
2017)

In the future self-driving technologies are forecasted to enhance driver safety,

reduce traffic congestion etc. (Thrun, 2010). The number of sensors

implemented in the automobile is constantly increasing (Manyika et al., 2013).

Infotainment systems are offering more entertainment options (Regan, 2004)

and connected automobiles do not only offer wireless internet access, but

also connection to a broad network of automobiles and enable V2V

communication. Gesture control systems are developing; augmented reality

is included into the automobile interface (Ropert, 2014). The amount of

18

software and technology built into automobile interfaces and applications for
navigation, music, internet, communication and connection to social networks
challenge the driver by providing a large volume of information. Drivers are
preoccupied with electronic devices, potentially degrading performance,
impacting road safety. Attention sharing in a hazardous environment, such as
the driving environment, presents a significant safety issue (Regan, 2004).

Self-driving automobiles could transform the driving environment from a
potentially stressful environment requiring high concentration and
responsibly, to an environment for relaxation and social interaction (Rand
Report, 2014). While increasing automation is predicted to enhance safety
and decrease the environmental impact of automobiles, concerns regarding
self-driving vehicles arise, such as fears of privacy issues and hacking of on-
board computer systems (Ropert, 2014). The greatest obstacle in the way of
developing technologies for the automotive environment is human
acceptance (Thrun, 2010). With increasing automation and growing
sophistication of in-car technologies customer behaviour and needs will
change (Gao et al., 2016) and a more emotional relationship between drivers
and their automobiles will be created (Michael, 2001; Hagel et al., 2017).

When comparing the developments in the automotive industry to Maslow’s
Pyramid of Human Needs multiple comparisons are obvious. In his book ‘A
Theory of Human Motivation’, Maslow explains human motivation through his
pyramid visualising the hierarchy human needs. Even though the model was
created in 1943, it is prescient in its content and often referred to in modern
research. In Maslow’s Pyramid of Human Needs the appearance of one need
rests on the satisfaction of the prior need (Maslow, 1943). This approach has
been used as an analogy in multiple disciplines, such as marketing or design
(Heppard, 2015; Baches, 2016) and has some strong similarities to the
developments in the automotive industry and research (Figure 2).

19

Figure 2 Analogy between Maslow’s Pyramid of Human Needs and historical developments in the automotive industry (based on
Akamatsu, Green and Bengler, 2013; Hagel et al., 2017; Miller, 2001; Normark and Gkouskos, 2012)

20

Figure 2 Analogy between Maslow’s Pyramid of Human Needs and historical developments in the automotive industry (based on Akamatsu,
Green and Bengler, 2013; Hagel et al., 2017; Miller, 2001; Normark and Gkouskos, 2012)

While the first four stages of physiological needs, safety, love/belonging and
esteem can be seen in parallel to the past developments in the automotive
industry of transportation, safety and human factors, comfort and experience,
the peak of the pyramid is based on predictions and recent and future
developments. In order for future automotive design to adapt to the
mentioned challenges the traditional automotive design process needs to
adapt (Giuliano, Germak and Giacomin, 2017). A shift towards human
emotion and a more human-centred approach to automotive design is
required. To accomplish this the significance of the role of emotion in the
driving environment must be gauged.

1.2 The Importance and Role of Emotion

The automotive environment is highly emotional (Sheller, 2004),
encompassing the driving experience, corresponding feelings and emotional
states (Carrabine and Longhurst, 2002), the social relationship between the
automobile and driver (Hagman, 2010), the emotional values an automobile
objectifies (Urry, 2004) and the identification of drivers with their vehicles
(Edensor, 2004). The emotional components of the driving experience
become apparent through expressions like “thrill of driving”, “passion of the
collector” (Sheller, 2004, p.5) and often directly involve expressions with a
high emotional value (Miller, 2001; Sheller, 2004). Semantics addressing
emotions, feelings or desires are used in marketing strategies of major
automobile companies such as BMW with the headline “Joy is BMW”
(Blanckenberg, 2009) or Peugeot using “Motion and Emotion” (Peugeot
Motion & Emotion, 2010). The emotional significance of the pleasure of
driving for instance can be explained by the “kinaesthetic experience” (Katz,
2001, p.144), which defines driving as an embodied and sensuous
experience and emphasises the feelings created through the experience of
“movement and being moved together” (Sheller, 2004, p.68).

The relationship between drivers and automobiles might be the most

21

emotional human-machine-interaction (Michael, 2001; Miller 2001). The
automobile is personified as a companion or even as wife or lover (Sheller,
2004). This relationship is reinforced by major automobile companies,
offering applications emphasising the connection between driver and
automobile such as watches or mobile applications, which create a 24-hour
bond between human and machine (e.g. BMW smart watch). Moreover, the
automobile is strongly connected to significant human values, such as
freedom and independence (Sheller, 2004). The ownership of an automobile
creates the feeling of power and freedom; the automobile is often referred to
as a status symbol eliciting feelings such as pride and power (Urry, 2004).
Moreover, ownership is strongly connected to the identification of a driver with
his/her vehicle (Hagman, 2010). The purchase choice is strongly influenced
by the values and characteristics the buyer can identify with (Hagman, 2010).
The decision to purchase an automobile is an emotional one, as the
identification with the vehicle includes self-reflection and self-realisation
(Carrabine and Longhurst, 2002). A strong identification again leads to a
stronger connection between human and automobile (Sheller, 2004).

“Car design influences the lives of millions of people throughout the
world. Whether the car serves as merely a practical means of
transport or as an extension of one’s personality, its design and
brand will always attract comment.” (Newbury, 2002, p.11)

Research (Hsu, Lu and Ho, 2013; Warell, 2008) into the visual appeal of
automobiles, such as the study by Cornet and Krieger (2005), identified the
four most important influences on the purchase of an automobile as “exterior
styling”, “interior styling”, “trendy” and “makes me feel attractive” (Cornet and
Krieger, 2005, p.1). Factors such as expression, emotional response and
brand impression are significant to the automotive industry (Cornet and
Krieger, 2005). Further studies covered the influence of the exterior styling
(Desmet, 2002), the social roles (Pelly, 1996) and other emotional aspects in
the automotive industry.

22

Negative emotions can have a major impact on driving performance, and
cause accidents or road deaths (Dement, 1997; Wells-Parker et al., 2002).
These emotional states include aggressiveness and anger (Wells-Parker et
al., 2002), fatigue and drowsiness (Lyznicki et al., 1998), stress (Matthews et
al., 1998), confusion (Ball and Rebok, 1994), nervousness (Li and Ji, 2005)
and sadness (Dula and Geller, 2003). Aggressiveness and anger can cause
aggressive driving behaviour (Wells-Parker et al., 2002). Fatigue and
drowsiness can influence concentration, judgement, perception and reaction
time, potentially causing accidents (Dement, 1997). Another common
problem is driver stress, often leading to a significant decrease in driving
performance (Hoch et al., 2005; Uchiyama et al., 2002). Driver confusion can
cause increased reaction time, or decreased decision-making ability
(Jonsson et al., 2005). Confusion is prevalent amongst elderly drivers, but
with the increasing sophistication of digital technologies it is expected to
become a greater issue in the future. Nervousness and sadness can
negatively influence the driving performance and lead to human error and
accidents (Li and Ji, 2005). Human error is still the largest cause for accidents
and road deaths (National Highway Traffic Safety Administration, 2008).
While the influence of negative emotions on driving has been studied in the
past, in-depth investigations of multiple emotional states during driving are
limited. This research takes the view that a systematic investigation of
multiple emotional states is necessary to respond to the aforementioned
challenges in automotive design.

Autonomous vehicles offer many potential benefits for the future, including
reduced CO2 emission and fuel consumption (Bullis, 2011), increased safety
and reduced fatalities (Manyika et al., 2013) and decreased congestion
(Dumaine, 2012). Autonomous driving could lead to social benefits;
occupants could decide freely how to use their time in the vehicle. Predictions

1.3 The Importance of Emotion for Autonomous
Vehicles

23

present the idea of the future automobile as a living room, workplace or the
new ‘third place’ (e.g. Mercedes concept self-driving automobile introduced
at CES 2015 Las Vegas).

Automotive technology is progressing strongly towards higher automation
(Manyika et al., 2013), with self-parking systems, steering assistance and
automatic breaking systems. Toyota, General Motors, Mercedes, BMW, Audi
and Volvo are testing autonomous systems (Manyika et al., 2013). Several
predictions see autonomous automobiles on the streets by 2025-2030 (Gott,
2014; Jacques, 2014; Stanley and Gyimesi, 2015).

“That would translate into approximately 10 to 20 percent of the
1.2 billion private automobiles projected to be on the road in 2025
having the ability to self-drive in at least half of all traffic situations.”
(Manyika et al., 2013, p.81)

The largest issue facing autonomous vehicles is therefore not the technology,
but regulatory frameworks and most importantly public support and human
acceptance (Manyika et al., 2013; Thrun 2010). Without human acceptance,
there is no foundation for connected, self-driving automobiles. Thrun states
that:

“We need to overcome the old belief that only people can drive cars,
and embrace new modes of transportation that utilize the twenty-first
century technology.” (Thrun, 2010 p 106)

Schoettle and Sivak conducted research into the acceptance of self-driving
vehicles and 43.5% of the participants stated that they always want to be in
full control of their vehicle, only 15.6% had a positive point of view towards
autonomous automobiles and more than 60% of participants expressed
concerns about riding an autonomous automobile (Schoettle and Sivak,
2014). With the automobile taking over tasks that the driver once used to
perform, issues of acceptance, trust and pride can arise (Koo et al., 2015,
p.2). To prevent this from happening Koo et al. emphasise the “need for

24

designers to observe this phenomenon from the perspective of the human
driver” (Koo et al., 2015, p.2). Recent surveys investigating the public opinion
of partially autonomous and self-driving vehicles reinforce the need for further
research and adaptations (Business Wire, 2014; Cisco, 2013; Schoettle and
Sivak 2014; Silberg et al., 2013). The role of emotion research becomes
highly significant when taking into account upcoming changes in the
automotive environment due to automation. Koo et al. state the concern that
“cars are starting to make decisions on our behalf” (Koo et al., 2015, p.1).

When looking at the stages of vehicle automation (Figure 3) it becomes
apparent that we are currently in the progress of entering a stage of highly
automated vehicles, facing full automation in 2025.

Figure 3 Arch of vehicle automation (Horrell, 2014)

For highly automated vehicles where the driver still has an active role and
control is shared between the automobile and the driver, the role of human-
automobile interaction is highly significant. While the role of the driver in a
fully manual automobile, as well as in a fully self-driving automobile is
obvious, the biggest challenges arise with partially autonomous cars
(Norman, 1990). With automobile and driver cooperating, interaction needs
to happen on an affective level to create a successful control loop. To keep
the human informed, the automobile must understand and respond to human
behaviour and emotions (Shaikh and Krishnan, 2013). Therefore, a high level
of understanding of drivers is required.

In the fully automated stage, the role of the driver will change into a passenger
role, with the automobile taking over all tasks which used to be human-

25

controlled. For the passengers to trust their self-driving automobile an
affective passenger-car communication is required. The automobile must
understand human needs and respond to behaviour and emotion; emotional
factors and affective states are therefore crucial for acceptance, safety and
comfort of future automotive design (Eyben et al., 2010). Furthermore, new
communication models between automobiles and pedestrians have to be
created in order to ensure the pedestrian’s safety and comfort. The
automobile will need to be able to predict the pedestrian’s behaviour (Meeder,
Bosina and Weidmann, 2017).

While multiple researchers are investigating the requirements stemming from
current and future stages of automation and the necessity to respond to
needs of drivers, passengers and pedestrians (Eyben et al., 2010; Meeder,
Bosina and Weidmann, 2017; Shaikh and Krishnan, 2013) further research
on methods of emotion investigation during driving is required.

To meet human requirements for coping with current and future automobile
technology, it is important to understand the multi-layered emotional role of
the automobile (Sheller, 2004). To achieve this, studies of human behaviour
and emotional states in the driving context are required. We need to be able
to predict the channels of user behaviour and emotion while driving. As
Norman states “we must design our technologies for the way people actually
behave, not the way we would like them to behave” (Norman, 2007, p.12).
Furthermore, a shift of the main focus of the automobile industry from human
performance to human behaviour is required for future design to address
issues such as the greater sophistication of technologies and the increasing
complexity of the automotive habitat. To respond to these challenges a better
understanding of human behaviour, needs, desires and emotions is
necessary (Giuliano, Germak and Giacomin, 2017).

1.4 Current Technology for Estimating Human Emotion
and Current Uses of the Collected Information

26

While most research approaches focus on negative emotions and their
consequences (Deffenbacher et al., 2002; Parkinson, 2001), the focus is
gradually shifting towards the investigation of multiple emotional states in the
automotive environment. One approach to a better understanding of human
emotion which has recently been applied in automotive research is affective
computing, the study of systems or devices which can recognise, interpret or
process human emotion (Picard, 2003). Different physiological measurement
tools (e.g. Galvanic Skin Response, Respiration Rate), behavioural sensors
(e.g. Facial-Expression Analysis, Speech Analysis) or self-report tools (e.g.
Car Logs, Self-Assessment Manikin) have been applied to driving research
(Healey, 2000; Hoch et al., 2005; Jones and Jonsson, 2005). Moreover,
numerous modern human–centred design approaches combining various
methods have been applied to automotive research and design, to investigate
the drivers’ and passengers’ behaviour, emotion and needs and improve the
driving experience (Gkatzidou, Giacomin and Skrypchuk, 2016; Giuliano,
Germak and Giacomin, 2017).

Current research and appropriate methods for the investigation of emotions
during driving are not only limited but also not comparable due to major
differences in applied methodologies, investigations of emotion and
measurement tools. Furthermore, research methodologies and results are
often highly influenced by their limitations (Chapter 3.2 Emotion
Measurement in Automobile Research). Therefore, further research
systematically investigating emotional responses within the automotive
habitat is required. This research uses identification of the limitations and
shortcomings of previous efforts to direct the line of investigation. Through
the development of specific queries that must be answered, an improved
framework for the study of human emotion in the automotive environment
may be developed.

27

1.5 Research Questions

This research takes the view that there is a need for the automotive industry
to shift the focus towards the driver’s behaviour, emotions and needs. A better
understanding of the driver’s emotional response to changes in the driving
environment is required, especially considering current and future
technological developments. The aim of this research is to investigate the
measurability, natures and causes of emotional responses during simulated
and on-road driving. Furthermore, the ability to causally assign individual
emotional responses to specific sources of the emotion will be explored.
Additionally, this thesis will investigate the assigned causes for emotional
events in the driving environment and the influence of different road types
and situations on those. Finally, a dataset of emotional responses and their
causes in different driving settings will be created. A comparison featuring a
statistical analysis of the collected results will be conducted to investigate how
different driving settings influence drivers’ emotions. Therefore, the main
research questions of this thesis and are:

How can human emotional responses be measured in an automotive
environment?

Can emotional responses be triggered and measured in a simulated
driving environment? What are the major challenges?

What are the typical natures, frequencies and causes of emotional events
in a natural real driving environment?

What are the typical natures, frequencies and causes of emotional events
in a partially controlled real driving environment?

How can the typical natures, frequencies and causes of emotional events
between two on-road studies in different settings be compared?