Algorithm for Ethical Decision Making at Times of Accidents for Autonomous Vehicles

摘要

Before autonomous cars can become widespread, car manufacturers must solve an important ethical dilemma of algorithmic morality. In time of unavoidable accidents autonomous vehicles face critical decision making situations where two or more lives are at stake. In times like these vehicles face ethical dilemmas i.e. whether to sacrifice lives for minimizing the number of deaths. In this article this programming ethic for self driving cars is visualized with an algorithm. A priority queue based min-heap sorting algorithm is used to categorize lives of people according to some given parameters. In this algorithm deletion, insertion etc. operations are used to prioritize among nodes with different values. We combine these operations using a BST as a priority queue. This algorithm also uses the term probability of survival, which indicates the probability of a person surviving. A flexible and fragile algorithm is presented so that customers and car manufacturers can design the ethical decisions based on their own ethics.