Ethics of Autonomous Vehicles

Mankind is driven by a preoccupation with technology.  Nowhere is this seen more than in the area of transportation. Railroads, airplanes, and automobiles have been a major driver of new technology and innovation–filling the pages of history with many historic firsts.  What was set in motion by the forefathers of invention–James Watt, the Wright Brothers, Gottlieb Daimler and Wilhelm Maybach–continues to propel our imaginations ever forward.  With unquantiable millions of transportation vehicles in operation today, one of the highest priority concerns relates to improving occupant safety and overall vehicle operating efficiency.  To help improve the quality of this transportation, mankind has proceeded by implementing mechanical, electrical, electro-optical, and more recently artificial intelligence (AI) technology into these vehicles.  Yet what are the ethical implications involved with a future where electronics and AI technology take over much or most of the responsible agency control of vehicles from humans in an attempt to have modes of transportation that are safer and more efficient because they are operating from an  autonomous technology system (ATS)? ^[1]  Such an AI will require a sufficiently defined technology platform that allows it to function relatively autonomously.  Part of this system will include how it makes ethical decisions in order to promote well-being and reduce and eliminate harm. 

Advanced Driver Assistance Systems (ADAS)

Fighter pilots today are surrounded with a full complement of computers and artificial intelligence that assists them with flying one of the highest performance vehicles man is presently capable of designing.  AI is largely in control of the particulars of the plane with the pilot being in the position of the ultimate and overriding control agent allowing him to adapt and take over control as needed.  In the future, the continuing development in electronics and AI technology portends for more common items such as automobiles to have many similarities regarding the overall agency control of the vehicle.

Presently, many electronic technology companies that support the automotive industry are squarely focused on addressing the technical needs of Advanced Driver Assistance Systems (ADAS), by developing both adaptive and predictive systems and components that will allow for better and safer driving.  ADAS assists the agents in charge of the vehicle by providing warnings or taking actions to reduce risk or through the automation of some portion of the control task of operating the vehicle so that safety and performance are improved.  The current state of ADAS is largely cooperative with the driver: that is, the human-to-machine interface in ADAS functions as part of the overall agent control of the vehicle with the human still maintaining the overall responsibility for the vehicle.  It is expected that over time, developments in technology will be successful in having more and more agency and control of the vehicle moved into the ADAS.  It is anticipated that ADAS may ultimately develop further into autonomous systems that will offer a superior level of intelligence and the ability to respond quicker and with greater beneficial results than when a human agent is the controlling driver. 

Electronic Components in ADAS

A variety of global electronic component leaders are focused on developing products that support ADAS and, by extension, technologies that will enable the future reality of autonomous vehicles.  ON Semiconductor provides robust, AEC-qualified, production part approval process (PPAP)-capable products for automotive applications such as the NCV78763 Power Ballast and Dual LED Driver for ADAS front headlights.  Freescale Semiconductor is helping to drive the world's most innovative ADAS solutions with their automotive, MCU, analog and sensors and digital networking portfolio expertise.  Their new FXTH8715 Tire Pressure Monitoring Sensors (TPMS) solution, which integrates an 8-bit microcontroller (MCU), pressure sensor, XZ-axis or Z-axis accelerometer and RF transmitter, was driven to market by the requirement for improved safety.  AVX, a technology leader in the manufacture of passive electronic components and interconnect products worldwide, has developed the VCAS & VGAS Series TransGuard® Automotive Multi-Layer Varistors (MLVs) which are well suited for protection against automotive-related transients in ADAS applications.  Delphi Connection Systems, a world leader in supplying vehicle electric and electronic distribution systems, is now offering the high performance APEX® Series Wire Connectors which are ideal for challenging automotive applications due to their robust design and reliability.  As the importance and impact of ADAS continues to expand, electronic component manufacturers will continue to develop new products to bring to market to ensure that vehicles can operate with greater reliance on highly dependable and efficient technology systems.

The Autonomous Vehicle

Improvements in electronics and AI technology come from the use of brain power to discover new products and less costly methods of production.  It is the dream of many that future technological advancements will allow for completely autonomous vehicles to dominate the roadways of the world.  These autonomous vehicles would be programmed to avoid accidents and damage to passengers as well as collateral damage, while simultaneously being programmed to optimize efficiencies in travel by optimizing various performance factors of the vehicle on-the-go so as to minimize expenses and maximize the functional utility of the vehicle.   

Ethics of Autonomous Vehicles

There is no doubt that everyone desires vehicles to operate safely and efficiently for all concerned.  In operating a vehicle upon today’s roadways, the laws currently on the books place the responsibility squarely on the human driver to ensure that everyone’s well-being, both inside the car and outside the car, are protected and that harm does not come to oneself or others. 

It is possible to conceive of an automobile as a tool whose primary function is to get people from Point A to Point B safely and efficiently.  However, there is little doubt that a vehicle can also be operated by an agent and used for vicious purposes, such as a murder weapon, where it is used to run down people as people attempt to cross the street. 

This is where a proper understanding of ethics must come into view.  Ethical judgments are invariably a tricky subject as they bring into play at least three different factors or perspectives including (a) a standard of judging or norm, (b) a contextual situation and environment, and (c) the relationship and impact relative to a person, both internally and externally.  Failure to take into account all of these various factors appropriately and correctly may lead to a reductionistic and non-optimal ethical judgment.  

If agency and responsibility are to become increasingly less dependent on the individual driver’s performance in operating the vehicle and increasingly more reliant on technologies such as ADAS and ultimately ATS located in vehicles, then how is society to handle the potential change in agent responsibility as it shifts from the individual operating the vehicle toward the vehicle itself?   The logical answer is the manufacturers of the autonomous vehicle will be seen as the liable agent as they are assembling and creating the decisive technological platform that controls the agency.  Naturally, vehicle manufacturers do not want to accept this risk liability unless the market provides them a proper economic incentive to do so, which at present does not exist, but may in the future.

Consider the following hypothetical situations.

1. You are the driver of a vehicle driving down the highway at 100km/hour with traffic around you and an animal jumps out in front of you.  You swerve to avoid hitting the animal but in doing so hit another car.  You had the choice and you made it.  You are responsible for the accident but you have saved an animal.  But what if the people you hit are significantly hurt? 



2. You are the driver of a vehicle driving down the highway at 100km/hour with traffic around you and a person appears from nowhere out in front of you.  You swerve to avoid hitting the person but in doing so hit another car.  You had the choice and you made it.  You are responsible for the accident but you have saved a life.  But what if the people you hit are significantly hurt?



3. You are sitting in an autonomously driven vehicle and are traveling down the highway at 100km/hour with traffic around you and a person appears from nowhere out in front of you.  Your car decides it cannot run over a person or swerve into another car at such high speeds without risking your life and the occupants of the other vehicle.  The car decelerates rapidly and swerves so that you hit a wall and are seriously injured.  You are not responsible for the accident but the autonomous vehicle is.  A human life has been saved by not being hit in the road.  The other vehicle emerged without incident because the autonomous vehicle was programmed not to harm others.  You are significantly injured with spinal cord damage.  You blame the situation and the autonomous program.

As one can see in just three scenarios, the situations and ethical dimensions that arise are extremely important and of consequence for all involved and were determined within split seconds.  How will vehicles be able to be programmed such that these situations do not become an endless series of litigation regarding responsibility? 

Some people question whether the world is ready for drivers to yield control of their vehicles to an autonomous vehicle.  To many, it may seem that hanging onto control of the vehicle is the best way to ensure that one is protected from an autonomous vehicle that is not programmed for such complicated decisions.

Perhaps the best response is simply to proceed with experimentation and caution.  Technology has a way of often solving problems in a manner that cannot be predicted or predetermined.  It may be that in the future autonomous vehicles can be technologically equipped in such a manner that the outcomes are always able to be best for the norms, contexts, and individual situations.

Technology will continue to progress and provides opportunities for mankind to reap the benefits.  Improved modes of transportation will continue to be developed helping mankind to move with greater freedom, flexibility, efficiency, and safety.  Electronic components and AI technology will play a significant role in these coming developments.  As things move forward, mankind must exercise wisdom and ensure that as more control of the operation of the vehicle is ceded over to the vehicle itself that the correct ethical structures and programs are in place to protect human life, society, and property so that autonomous vehicles turn out to be a blessing for everyone.

References:

1. Autonomous Technology Systems (ATS) are for the purposes of this article considered to be any technology system (likely consisting of very significant and complex electronic and AI technology interrelationships) that have been programmed by its creator to select from among relevant available possibilities (it has the ability to choose) by a decision making process relative to the situations that it faces such that the outcomes of events are determined by its response.  This means that the ATS is the responsible agent because it has all the relevant power to act. 

About the Author

Paul Golata is a Senior Technical Content Specialist at Mouser Electronics where he is tasked with providing design engineers the latest information and technical content. Prior to Mouser Electronics, he served in various roles for Hughes Aircraft Company, Melles Griot, Piper Jaffray, Balzers Optics, JDSU, and Arrow Electronics. Mr. Golata holds several degrees including a BSEET, MBA and a MDiv.

About Mouser Electronics