Human Driver’s Reasoning on Moral Dilemma of Autonomous Vehicles: Values and Themes

Abstract: Autonomous Vehicles (AVs) can make optimized decisions with its advanced sensors, algorithms, and control even during accidents. The computing capacity of AVs allows the public to rediscover the classic trolley dilemma in the modern context. Consequently, the moral dilemma of AVs has been questioned in preparation of wide use of AVs. This study starts with a view that consequentialism alone cannot be accepted as social morality in the moral dilemma situations. Instead, we focus on the process through which humans choose ethical values and themes. First, “when can AVs face a moral dilemma in real-life traffic accident situations?” Second, “what factors can be considered in AVs moral dilemmas from the driver’s perspective?” Third, “how do the factors play a role in driver’s moral reasoning? How can we understand and explain the reasoning process?” While answering the questions, we aim to present an in-depth explanation of driver’s moral reasoning for AV moral dilemma situations. Moral dilemma vignettes for AVs were designed based on real crash data and in-depth interviews with drivers. With the vignettes, a thought experiment with 33 participants was conducted; think aloud method and open-ended interviews were used to examine participants’ reasoning processes. The data were analyzed using the thematic analysis leveraging graph representation method. Then we discuss the themes of driver’s moral reasoning. We found the codes of driver’s moral values by focusing on the ideas that appear in personal moral dilemmas. The codes were referred to normative ideas, procedural ideas, and actions. Also, we present three themes of driver’s moral reasoning. Theme I, Safety-Oriented Reasoning was derived by driver’s responsibility and empathy on oneself, and was faithful to the safety of oneself and passengers. Theme II, Justice-Oriented Reasoning focused on norms such as minimization of casualties (utilitarian value) and responsibility for fault. Theme III, Situational Reasoning led to different decisions depending on contexts, showing the sub-themes of the other two themes and great impact of guilt. A variety of codes and themes shows the diversity of human driver’s morals in unavoidable vehicle crash situations. Although our experiment was restricted to the Korean population, the result can be extended to cultural comparisons by providing the case of Eastern collectivist culture.The result provides qualitative insight to integrate human morality and intelligent systems. As well as, the results can be used for a qualitative tool for analyzing AVs algorithms of judgment in an emergency by comparing them with human driver’s moral reasoning. Given the difficulty in the cooperation of ethicists and engineers in manufacturing AVs, our results can serve as a communication tool for discussion for developing the algorithms and conveying moral perspectives of lay drivers, who are the potential users of AVs. The opening of such discussion among ethicists and engineers, or other stakeholders including potential users, will benefit social consensus for AV ethics. The results are also expected to help engineers think about the difference between AV-human judgments in the moral dilemma situations, as well as get insights for their algorithms. AV engineers can use our vignettes as representative scenarios of the moral dilemma in real traffic.

Keywords: Moral Dilemma, Autonomous Vehicle, Moral Reasoning, Thematic Analysis

DOI: 10.54941/ahfe1001053

Cite this paper:

​