• NAKANO Kimihiko
  Professor, Institute of Industrial Science, The University of Tokyo

From FY2020 to FY2023, we examined ELSI themes in the social implementation of autonomous vehicle driving through a project titled “Development of a Social Implementation Method for Autonomous Driving Technology Based on ELSI” under the ELSI Program. The fundamental question, which is also used as the title of the project, is: “Can People and Society Accept the Mistakes Made by Machines?” Ensuring the safety of autonomous vehicles is a prerequisite for the social implementation of autonomous driving. However, unfortunately, the risk of accidents cannot be eliminated entirely. Nobody wants to cause an accident. This poses a formidable challenge for engineers seeking to expand the industrial applications of science.

Defining the necessary scope of safety for social implementation can be a challenging task. One approach is to statistically demonstrate that autonomous vehicles are safer than human-driven vehicles by showing that, based on historical data, the number of accidents per unit distance is lower for autonomous vehicles than for manually driven vehicles. However, accident victims are not concerned with the frequency of other accidents, but with receiving compensation for the damage they have suffered. Furthermore, it would be unacceptable if there were no one to take responsibility when personal injuries occur. Researchers affiliated with Meiji University have been discussing compensation for accident damage as part of this project. We expect to see institutional designs on this issue in the future. However, the question of responsibility in the event of an accident must be addressed not only as a legal issue, but also as an ethical and social issue. While a reduction in the number of accidents can be seen as an indicator of the safety of autonomous driving, when humans assign tasks to machines, they generally expect the machines to perform them flawlessly. Relying solely on statistical data to discuss safety would be ethically inadequate and may not convince people.

As the elderly population grows, there is a growing recognition that reliance on private automobiles for mobility is reaching its limits. At the same time, factors such as declining economic performance and a shortage of drivers are making it difficult to sustain public transportation. Against this backdrop, expectations for the social implementation of autonomous driving are rising. However, it appears that these expectations may be outpacing the current reality. If the expectations for autonomous driving technology become overly optimistic, and this is introduced into society without a full understanding of the risks involved, it may become excluded by the society once an accident occurs. It is crucial to have a proper understanding of the current status of autonomous driving technology. Therefore, in this project we have emphasized activities that prioritize interactive scientific communication.

The author has pursued answers to fundamental questions through his involvement in research and development aimed at the social implementation of autonomous driving. One approach to addressing this issue is to identify and list the anticipated risks associated with the use of autonomous cars and proactively plan how to manage them. The amount of work involved is considerable, and there is no way to guarantee that all risks have been identified. In the end, it is impossible to eliminate the risk of accidents altogether. When an accident unfortunately occurs, all we can do is analyze the cause, improve the system, and through this iterative process, incrementally improve safety with the expectation of wider societal adoption over time.

The greatest possible effort should be made to ensure that safety is given the highest priority. In addition, it is necessary to clarify where responsibility lies in the event of an accident. However, in order for society to fully benefit from the innovations brought about by the societal implementation of emerging science and technology, it is essential that any accidents that occur be thoroughly analyzed and investigated, and that measures be taken to prevent similar incidents in the future. For more information, see “Proposal on the Status of Accident Investigation System”*1 (June 23, 2005) by the “Committee on Safety Engineering, Committee on Human and Engineering Sciences” of the Science Council of Japan. The Proposal emphasizes the need for an independent organization to conduct accident investigations in order to prevent similar incidents and improve safety measures. When an accident occurs, it receives widespread media attention and public interest. However, by the time the investigation is complete and preventive measures are implemented, public interest often diminishes. While taking responsibility for mistakes and providing financial compensation is necessary, more attention should be given to preventing future occurrences and improving safety.

As new technologies are introduced into society, it is critical to have transparent discussions to ensure a proper understanding of both the benefits to society and the negative aspects, including the fact that risks cannot be completely eliminated. I believe that society can accept new technologies that are not risk-free, provided that continuous efforts are made to ensure safety, and that when accidents do occur, a cycle of activities is established to compensate victims, implement preventive measures, and improve safety.

*1Proposal on the Status of Accident Investigation System by the Committee on Safety Engineering, Committee on Human and Engineering Sciences

Essay All Discours