ISY1000 Report: Professional and Ethical Practice in Self-Driving Cars

Verified

Added on 2022/09/28

AI Summary

This report delves into the ethical dimensions of self-driving cars, addressing key concerns such as liability in accidents, the potential impact on the insurance industry, and the pros and cons of implementing standardized artificial intelligence algorithms across automobile manufacturers. The analysis explores the levels of automation and their implications for assigning responsibility in the event of an accident, examining whether liability falls on the driver, the manufacturer, or other parties. Furthermore, the report investigates how the widespread adoption of self-driving cars could reshape the insurance landscape, potentially leading to changes in premiums and coverage. The report also evaluates the advantages and disadvantages of using standardized AI algorithms, considering factors such as ease of integration, potential for errors, and the impact on manufacturers' ability to differentiate their products. The report also examines the challenges of deep learning in the automotive industry, including the need for powerful hardware and solutions for sensor fusion. The report concludes with recommendations for ethical considerations in self-driving car software development.

Contribute Materials

Your contribution can guide someone’s learning journey. Share your documents today.

Running head: PROFESSIONAL AND ETHICAL PRACTICE 1
Professional and Ethical Practice
Student’s Name
Institutional Affiliation

Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.

PROFESSIONAL AND ETHICAL PRACTICE 2
Executive Summary
The development of self-driving car is on the increase with the aim to solve the problem of high
number of accidents experienced on the roads. Therefore, autonomous vehicles are being tested
to determine if they will provide a suitable solution to the accident menace. However, AVs are
faced by several ethical issues where they have caused accidents while undertaking test on the
road. This paper discuss various issues related to self-driving cars such as part liable in case self-
drive car is involved in an accident, impact of self-drive cars to the insurance industry and pros
and cons of AI algorithms as standard across automobile manufacturers.
Keywords: Self-driving cars, autonomous vehicles, artificial intelligence algorithm,
human driver and driverless cars.

PROFESSIONAL AND ETHICAL PRACTICE 3
Table of Contents
Introduction.................................................................................................................................................4
Where does liability reside when self-driving cars are involved an accident?.............................................5
How might the development of self-driving cars affect the insurance industry?.........................................6
Pros and cons of implementing a standardized artificial intelligence algorithms across automobile
manufacturers..............................................................................................................................................7
Pros of AI algorithms as standard across automobile manufacturers.......................................................7
Cons of AI algorithms as standard across automobile manufacturers......................................................8
Conclusion................................................................................................................................................. 10
Recommendations..................................................................................................................................... 10
References.................................................................................................................................................11

PROFESSIONAL AND ETHICAL PRACTICE 4
Ethical Decisions in Software Development: How Safe are Self-Driving Cars
Introduction
A report by the National Highway Traffic Safety Administration (NHTSA) note that most
car accidents in the roads is as a result of speeding, distracted drivers and drunk driving. In this
regard human errors which responsible for more than 90% of the car accidents could be
essentially eradicated with self-driving cars (Ramsey, 2015). Certainly, this leaves the remaining
prospect of accidents to autonomous vehicle design quality. Self-driving cars are skilled to make
decision in regards to when to break as well as where to steer using particular artificial
intelligence which is committed to achieving a narrow task (Goodall, 2014). Therefore, self-
drive cars are designed with various types of cameras, sensors as well as distance-measuring
lasers that provide information to a chief computer. The central computer then utilize artificial
intelligence to analyze the provided inputs and then make a decision. Even though self-driving
cars are regarded as the most suitable solution to eliminating accidents on the roads, they are also
not a hundred percent accurate.
The testing process of autonomous vehicles has been associated with several accidents.
For instance, in 2016, one of Google’s self-driving cars hit a bus while on a test drive in
California. The accidents happened because the self-drive car made an inappropriate assumption
regarding how the bus would react a particular situation. Before the self-drive car had identified
an obstruction on the road ahead and decided to stop waiting for the lane to clear and then marge
the other lane. Despite the fact that the self-drive car had detected a city bus approaching in the
lane, it made an incorrect assumption that the bus driver would slow down and that was not the

Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.

PROFESSIONAL AND ETHICAL PRACTICE 5
case. The car driver assumed the car would stay put and kept moving forward, thus the car pulled
out and hit the bus. In this respect accidents involving self-driving cars are normally as a result of
sensor error. In the contemporary world of today self-driving vehicles have been equipped with
complex sensing capabilities hence they are able to differentiate pedestrians from other objects
on the road such as road signs. Accordingly, self-driving vehicles are being developed to avert
accidents and to minimize the speed at impact if it is unavoidable (Borenstein, Herkert, & Miller,
2019). However, similar to humans self-drive vehicles are not capable of making moral decisions
prior to an unavoidable accident. The significance with self-driving cars it that they will be safer
as compared to humanoid drivers because they are more observant and can respond quickly and
make use of the braking system abilities to the fullest in an accident scenario. However, the main
ethical encounters facing self-driving car engineers is to determine when there is enough
evidence of safe behavior from simulations as well as organized on-road testing to introduce self-
driving vehicles on the road. At the end of the day, self-driving automobiles will be safer
compared to human drivers. This paper provide a response to the case study: Ethical Decisions
in Software Development: How safe are Self-Driving Cars under the following subheading.
Lastly, the paper concludes and provide commendation regarding ethical decisions to consider in
self-driving cars software development.
Where does liability reside when self-driving cars are involved an accident?
In case a self-driving car is involved in an accident where the liability reside will depend
on the automation level of the car that is involved in the accident. There are several levels of
automation for self-driving cars. These levels include: level one-this car has the ability to handle
one automatic task at a time such as automatic braking. Level two-this autonomous car has at
least two automated functions. Level three-this car is capable of handling dynamic driving tasks

PROFESSIONAL AND ETHICAL PRACTICE 6
but it still need the intervention of humans. Level 4-this care is completely driverless in some
environments. Level five-these vehicles have the ability to operate on their own in minus the
presence of a driver. Therefore, if the autonomous car involved in an accident fall under the first
three automation levels the liability will likely to fall on the driver’s because apart from the
vehicle being in position to automatically perform certain tasks, the presence of the driver is to
intervene. Hence the driver will be liable because it means the accidents could have been caused
by their fault (Fleetwood, 2017). On the other hand, in case an autonomous car that is involved in
an accident falls under level four or five the liability could most probably fall on the side of the
automobile manufacturer since the accident does not engage any input by a driver at all. In this
respect, the self-driving car is assumed to have gotten into the accident due to a failure on its
systems or a glitch on its sensors. As a result, this makes the automobile manufacturer to be held
accountable for the damages in case of an accident (Mackie, 2018).
How might the development of self-driving cars affect the insurance industry?
The development of self-driving cars when it reaches level four and level five which have
full autonomous capabilities which means no human will be involved it is lily to dynamically
change the insurance industry. In this regard, self-driving cars will pose a threat to the existence
the insurance industry. For instance automakers will assume liability, drop in insurance
premiums, and struggle in the insurance industry due to a decline in the number of drivers’
coverage.
Liability will be assumed by automakers: In case an accident occurs without the involvement of
a human driver, the manufacturer is the responsible for the damages. Thus, it means that the
vehicle manufacturer will assume the expenses of insurance.

PROFESSIONAL AND ETHICAL PRACTICE 7
Drop in insurance premiums because accidents will become rare: Experts have predicted that
with self-driving vehicles becoming the novel norm, it will result in a drop in accidents. Even
though, in the initial stages there could be challenges because drivers and driverless vehicles mix
up on the road because neither artificial intelligence drivers nor human drivers could be capable
of accurately foreseeing what to anticipate next from each other. Nevertheless, in the long run,
there is greater probability of a decline in the trend in accidents (Ramsey, 2015). On the same
note, the insurance industry will be forced to provide premium discounts, however with the
increase in safety being apparent through the deployment of level four and level five autonomous
vehicles, insurers will have to lower their premiums or risk being undercut by rivals.
Struggle by the insurance industry because more drivers will reduce or drop coverage: With
the advancements in autonomous vehicles, it will reach a point where drivers will not realize the
need to carry personal accident insurance.
Pros and cons of implementing a standardized artificial intelligence algorithms across
automobile manufacturers.
Pros of AI algorithms as standard across automobile manufacturers
There has a lot of developments taking place in machine learning algorithms because of
deep learning as well as with the increase in AI as a common platform. Therefore, because the
automobile sectors is set to perform several test changes in the upcoming years, a standardized
AI algorithms would make it easier to integrate various issues in the automobile industry. With
the rise in different issues which are currently being considered in the manufacturing process in
regard to artificial intelligence, automobiles are becoming more sophisticated, and integrated
systems. For example, Scheutz (2016) argue that if artificial intelligence could have existed fifty
years ago the automotive industry could be very far right now. As a result of increased deep

Paraphrase This Document

Need a fresh take? Get an instant paraphrase of this document with our AI Paraphraser

PROFESSIONAL AND ETHICAL PRACTICE 8
learning as well as greater computing power in addition to the existence of gigantic volumes of
data in the cloud it has facilitated the rapid growth of the present technologies. Indeed, all these
has been made possible because of standards. Recent research indicate that the rate at which
artificial intelligence based systems is being installed in new automobiles was approximately 8%
in 2015 Bringsjord and Sen (2016), but it is estimated that come 2025 the number would have
increased to more than 100%. This massive increase supports the usage of artificial intelligence
systems in cars could have to standard across the autonomous vehicle applications.
Deep learning is the main application in the automotive industry that comprises computer
perception and vision. The visual tasks include lane detection, recognizing road signs, pedestrian
detection as well as blind-spot monitoring can only work well with deep learning. Deep learning
is a more powerful and sophisticated computing operation which contains numerous neural
networks that contain a lot of hidden layers that add features and exceeds human coding
capability. In this sense, there is need for standard AI algorithms in the vehicle manufacturing
industry to allow for a system that lower tolerance to mistakes and capability to easily isolate
errors. At the same time this will allow the automobile industry the ability to manage
unpredictable situations in effectively.
Cons of AI algorithms as standard across automobile manufacturers
Deep learning faces a number of criticism. To begin with dee learning generalizes
scenarios at the time of training which is doubtful hence calls for through investigation. Standard
AI systems have a problem with extremely solid geometrical basis that theoretically can be
resolved more efficiently through other approaches (Herman & Ismail, 2017). Deep learning
calls for very powerful automotive grade hardware which could not be affordable for massive
production. On the same note, deep learning on data that comes from sensor fusion has not yet

PROFESSIONAL AND ETHICAL PRACTICE 9
been solved hence it pose a great challenge because of the diversified nature of the data and its
sheer volume.
Reaction of manufacturers to standardization of automobile AI algorithms
The automotive manufacturers have a different reaction towards standard AI algorithms
where it is argued that the automotive market to stand out they need solid working relationship
with OEMs, product differentiation built on security and safety features and collaborations with
other players across the value chain. In regard to solid relationships with OEMs automobile
manufacturers are ready to take proactive roles in advanced drive-assistance systems (ADAS)
development. Nyholm & Smids (2016) argue that if there is need to drive ADAS development to
allow OEMs the freedom to choose the most appropriate subsystems as well as control and
sensors systems from different tier-one vendors instead of depending on one source. With the
automotive industry taking charge of ADAS development it will ensure that OEMs has the
opportunity to differentiate their selves from rivals for autonomous-driving and driver support
functions.
Automated driving systems range from complete driver control (level 0) to full autonomy
(level 5). Should the degree of care exercised in developing vehicle software increase as the
level of autonomy increases, or should all vehicle software be treated with the same level of
care? Explain your answer.
No. All vehicle software should not be treated with the same level of care because they
these automotive levels are develop to realize a specific objective in the autonomous
development lifecycle (Casey & Niblett, 2016). For instance level 0 are manually controlled
where a human driver offers a dynamic driving task despite the presence of systems in place to

PROFESSIONAL AND ETHICAL PRACTICE 10
provide assistance. On the other hand, level 5 vehicle does not require human attention since the
dynamic task has been removed. This car does not need a steering wheel or breaking pedals.
Hence autonomous cars cannot have the same level of care because they have been designed to
meet different needs and user requirements.
Conclusion
In conclusion, the future of autonomous cars is exciting and promising because these
vehicles have the ability to pay attention unlike human drivers who can be distracted or drive
under influence of alcohol and other drugs leading to accidents. Consequently, there is need for
further studies in regard to development of automotive vehicles putting into consideration ethical
values.
Recommendations
The primary aim of developing autonomous vehicles is to ensure that the number of
accidents on the roads are reduced and even eliminated. In this light the paper provide the
following commendation in regards to ethical issues facing AVs.
The developers of self-driving developers should include both mandatory and personal
ethics settings which allows the user to choose the ethical preferences s for their vehicle to help
solve the responsibility problem.

Secure Best Marks with AI Grader

Need help grading? Try our AI Grader for instant feedback on your assignments.

PROFESSIONAL AND ETHICAL PRACTICE 11
References
Borenstein, J., Herkert, J. R., & Miller, K. W. (2019). Self-driving cars and engineering ethics:
the need for a system level analysis. Science and engineering ethics, 25(2), 383-398.
https://link.springer.com/article/10.1007/s11948-017-0006-0
Bringsjord, S., & Sen, A. (2016). On creative self-driving cars: Hire the computational logicians,
fast. Applied Artificial Intelligence, 30(8), 758-786.
Casey, A. J., & Niblett, A. (2016). Self-driving laws. University of Toronto Law Journal, 66(4),
429-442.
Fleetwood, J. (2017). Public health, ethics, and autonomous vehicles. American journal of public
health, 107(4), 532-537.
Goodall, N. J. (2014). Machine ethics and automated vehicles. In Road vehicle automation (pp.
93-102). Springer, Cham.
Herman, S., & Ismail, K. (2017). Single Camera Object Detection for Self-Driving Vehicle: A
Review. Journal of the Society of Automotive Engineers Malaysia, 1(3).
Mackie, T. (2018). Proving liability for highly and fully automated vehicle accidents in
Australia. Computer Law & Security Review, 34(6), 1314-1332. Mackie, T. (2018).
Proving liability for highly and fully automated vehicle accidents in Australia. Computer
Law & Security Review, 34(6), 1314-1332.
https://www.sciencedirect.com/science/article/pii/S026736491830356X
Nyholm, S., & Smids, J. (2016). The ethics of accident-algorithms for self-driving cars: An
applied trolley problem? Ethical theory and moral practice, 19(5), 1275-1289.
Ramsey, M. (2015). Self-driving cars could cut down on accidents, study says. Report predicts
mass adoption of auto-piloted vehicles beginning in about 15 years. The Wall Street
Journal, 5.

PROFESSIONAL AND ETHICAL PRACTICE 12
Scheutz, M. (2016). The need for moral competency in autonomous agent architectures.
In Fundamental issues of artificial intelligence (pp. 517-527). Springer, Cham.