Professional Documents
Culture Documents
Vishal Paper
Vishal Paper
Submitted on 09.03.2022
Error! Use the Home tab to apply Titel to the text that you want to appear here.
Abstract
The main objective of the launching of autonomous vehicles was to increase the efficiency of
the road infrastructure, improve the freedom of the drivers, and above all, reduce traffic acci-
dents. Despite the improvement in technologies, there might be situations where the auto-
mated vehicle needs to choose the most moral choice among the two evils. Binding regula-
tions must be made with the public interest and try to reduce the delay in the adoption of
AVs so that lives saved by AVs can be greater than the deaths caused in the case of delay-
ing AVs. The present study has tried to analyze how people would want the AV to act upon
these critical situations. In the second part of the study, the sample was again divided based
on their usage of the road infrastructure i.e. Drivers, pedestrians, and people with disabilities,
and asked which ethical principles concern them the most upon the introduction of AVs. The
main principles that were considered are responsibility, human autonomy, privacy, wellbeing,
and social justice. It was observed that people would prefer the utilitarian course of action
where the casualties are minimum. In ethical principles, responsibility was the biggest con-
cern among drivers and pedestrians which was followed by wellbeing, autonomy. The
drivers were worried about their data where the question of privacy arises and social justice
a concern by the people who couldn’t afford to spend a huge amount on an AV.
1. Introduction......................................................................................................... 3
1.1. Risks due to automated driving......................................................................3
1.2. Classification of the ethical principles and vehicle classes.............................3
2. Literature Review................................................................................................4
3. Methodology........................................................................................................5
4. Results................................................................................................................. 6
5. Discussion...........................................................................................................8
6. Conclusion...........................................................................................................9
References................................................................................................................. 9
Figure 1: Two traffic situations involving imminent unavoidable harm. The car must decide between
(to the left) killing one pedestrian or the passenger and (to the right) killing several pedestrians or the
passenger (Source: Bonnefon et al., 2016)
“AVs have the potential to reduce accidents but not all crashes will be avoided, some
crashes will require the AV to make ethical decisions to choose among two evils” (Bonnefon
et al., 2016). For example, as shown in figure 1, the first case shows if it’s moral to kill a pas -
senger or sacrifice the life of a passenger and the second case depicts if it’s moral to sacri-
fice the life of the passenger in order save that of many passengers. According to a study by
Bonnefon et al., 2016, it was found that people preferred utilitarian morals, according to
which the moral course of action is to minimize casualties. But AVs programmed to follow
this might discourage the buyers.
For a better understanding of people’s opinions on the extent to which they want their
vehicle to be automated, we have adopted the levels of vehicle automation according to the
Society of Automotive Engineers International (SAE, 2018) taxonomy. In level 1 (driver
assistance) and level 2 (partial driving automation), the human driver monitors the driving
environment and is assisted by a driving automation system for execution of either the lateral
or longitudinal motion control (level 1) or both motion controls (level 2). In level 3 (conditional
driving automation), an automated driving system performs all dynamic tasks of driving
(monitoring of the environment and motion control), but the human driver is expected to be
available for occasional control of the vehicle. In level 4 (high driving automation) and level 5
(full driving automation), an automated driving system performs all dynamic tasks of driving,
without any human intervention at any time. In level 4, the automated driving system controls
the vehicle within a prescribed operational domain (e.g. high-speed freeway cruising, closed
campus shuttle). In level 5, the automated driving system can operate the vehicle under all
on-road conditions with no design-based restrictions (Milakis et al., 2017).
2. Literature Review
“People let their intuitions and socio constructs such as trust, value congruency, emotions af-
fect their judgment, instead of crunching numbers like experts do” (Dogan et al., 2021). As
Roeser (2006 states in his study, technological risks are differently viewed by different pro-
fessions, for example, engineers define risk as a function of probabilities and unwanted con-
sequences such as the number of deaths, degree of pollution. Policymakers use cost-benefit
analysis to weigh the possible advantages of technology against its possible disadvantages
(Roeser, 2006). Examples of technological risks that spark heated and emotional debates
are cloning, GM – foods, and nuclear energy.
According to the study conducted by Montemerlo, 2008, the first benchmark test for
autonomous driving was done in realistic urban environments (Montemerlo, 2008). Spieser
et al, 2014, concluded in their paper that “automated vehicles (AVs) have the potential to in-
crease traffic efficiency, pollution reduction, and eliminate up to 90% of traffic accidents
Dogan et al, 2021 and their team have studied the ethical issues raised due to the deploy-
ment of AVs with drivers, pedestrians, and road users as the main focus groups. Their study
revealed that “In case of transgression traffic i.e. to comply with prohibited activities such as
crossing a continuous lane line to overtake a stranded vehicle, drivers considered it permiss-
ible while nondrivers considered it as impermissible”. They concluded that with the deploy-
ment of AV, there is strong support for a separate lane for the AV for the sake of safety and
efficiency (Dogan et al., 2021).
3. Methodology
To achieve the objective of the study, an empirical survey was conducted among different
age groups and compared between the region of Hyderabad, India, and Munich, Germany.
In the survey, we defined the scenarios which collected opinions of the people and how their
age. The following scenarios were depicted in the survey:
a) Kill one pedestrian crossing the road or sacrifice the life of the passenger
b) Kill several pedestrians crossing the road or sacrifice the life of the passenger
The same sample was further asked if they would be comfortable buying an AV if they were
programmed in such a way to sacrifice the life of the passenger and save the pedestrians.
On the next step, the participants were asked their opinion on the ethical principles:
c) Well Being – Reduction in travel time, improved road safety, environmental well be-
ing
e) Social Justice – How far can the low-income sectors afford to buy an AV
The above responses were divided according to their age as well as the user group of the in-
frastructure i.e. drivers and pedestrians.
With the results, we tried to build the relationship between the age and acceptance level of
an AV.
4. Results
Responses from 56 people in Germany were collected and the analysis is represented in the
following bar charts and line graphs.
Figure 2 shows the responses of the participants on a scale of 1 to 5 (1 being the choice to
protect the driver at all costs and 5 being the choice to protect the pedestrians whether it’s
moral to save the passenger or the pedestrian in two types of scenarios.
In scenario 1 where the car must decide between killing one pedestrian or killing its passen-
ger, people had a mixed opinion. Some people argued that since the count on the cases is
the same they couldn’t decide while some called it the fate of the pedestrian and the others
questioned why an innocent pedestrian should be killed. In scenario 2 majority felt that it
Was moral to sacrifice the life of the passenger to save that of several pedestrians so that
we could reduce the casualties.
20 30
25
15
20
10 15
10
5
5
0 0
1 2 3 4 5 1 2 3 4 5
Figure 2: Participants’ response if it’s moral to (to the left)) kill one pedestrian or its own passenger
and, (to the right) kill several pedestrians or its passenger.
Figure 3 depicts the automation level which the public would prefer according to their age.
The majority of the people are willing to accept automation level 4 where the system controls
the vehicle within the prescribed operational domain in the presence of a driver. It can also
30
25
20
15
10
0
1 2 3 4 5
Figure 3: Participants’ willingness on the automation level. The x-axis describes the level and the y
axis describes the no. of people. The legend indicates the age group.
The analysis of people’s concern of the ethical principles raised by AVs is represented in a
tabular form as shown in Table 1. A short overview of the arguments made by the parti-
cipants is also summarised.
Responsibility was the biggest concern among all the groups. In case of a death, drivers
were worried they might have to live with the guilt. Some of them also added that the guilt is
even more than that of being sent to jail.
Human autonomy was also a concern among all the groups as this would increase the de-
pendency on the technology which will create difficulties in the future if major changes hap-
pen. Others emphasized that they will gain freedom.
Wellbeing being the next concern, drivers felt AVs would increase the capacity and showed
reduction in travel time while pedestrians stated that it would stressful while interacting with
the vehicles like crossing the street.
5. Discussion
Although our study suggests that people would rather save the lives of pedestrians than
save themselves, most of them would prefer to buy an AV that would save their lives. This
suggests that some people might choose self-protective AVs and others will choose utilit-
arian ones. Binding regulations may provide a solution but the regulators themselves will
face problems with the disapproval of regulation that would enforce utilitarian AVs, and the
introduction of regulation could delay the adoption of AVs which would result in saved lives
by AVs outnumbered by the deaths caused due to manual vehicles.
Responsibility was the main concern by all the road user groups. Many stakeholders will be
involved in inducing AVs which makes answering the question of responsibility difficult.
Helen N., 1996 stated the accountability issues in a computerized society and defined this
phenomenon as a “problem of many hands”.
We also saw that drivers are more concerned about privacy and autonomy compared to the
others. They might be concerned about their travel records such as routes being taken and
the destinations they are visiting.
Human autonomy and well being were considered to be a concern of all the groups as from
pedestrians’ point of view, they should feel safe to cross a road even though there isn’t any-
one assisting the vehicle in driving and from the driver’s point of view they should be com-
fortable with the thought of giving full authority to the vehicle in driving.
Social Justice is a big concern not only for the people from the lower-income class but also
for the people who need to make some customized changes to the vehicle due to their phys-
ical problems such as disabilities. This category of people might have to pay more for the
AVs which are already having high initial costs.
According to the study by Maurya (2022), the majority of the people are willing to accept
automation level 3 which is conditional driving automation in which the driving system per-
forms all the dynamic tasks but the human driver is expected to stay. While in Germany
people are ready to accept up to high driving automation (Level4). This difference might be
because of the existing infrastructure of India as it is still a developing country unlike Ger-
many, the mixed traffic conditions, and the extent to which the people could afford it.
The ethical principles that were stated in the paper were presented to different road user
groups. Although there were some deep concerns about the impacts it would have on an in-
dividual or a society, there are many potential benefits associated with it provided it has to
meet the expectations of the people.
References
Bonnefon, J. F., Shariff, A., & Rahwan, I. (2016). The social dilemma of autonomous vehicles.
Science, 352(6293), 1573–1576. https://doi.org/10.1126/science.aaf2654
Dogan, E., Barbier, C., & Peyrard, E. (2021, April 26). Public perception of ethical issues
concerning automated mobility: A focus group study among three road user categories.
ACM International Conference Proceeding Series. https://doi.org/10.1145/3452853.3452877
Milakis, D., van Arem, B., & van Wee, B. (2017). Policy and society related implications of
automated driving: A review of literature and directions for future research. Journal of
Intelligent Transportation Systems: Technology, Planning, and Operations, 21(4), 324–348.
https://doi.org/10.1080/15472450.2017.1291351
Roeser, S. (2006). The role of emotions in judging the moral acceptability of risks. Safety
Science, 44(8), 689–700. https://doi.org/10.1016/j.ssci.2006.02.001
Spieser, C., Kevin, K., Ballantyne, R., Zhang, E., Frazzoli, D., Morton, M. P., Spieser, K.,
Treleaven, K., Zhang, R., Frazzoli, E., Morton, D., & Pavone, M. (2014). Road Vehicle
Automation. In Lecture Notes in Mobility. Springer.
http://hdl.handle.net/1721.1/82904http://creativecommons.org/licenses/by-nc-sa/3.0/