Machine Ethics and Automated Vehicles

Road vehicle travel at a reasonable speed involves some risk, even when using computer-controlled driving with failure-free hardware and perfect sensing. A fully-automated vehicle must continuously decide how to allocate this risk without a human driver’s oversight. These are ethical decisions, particularly in instances where an automated vehicle cannot avoid crashing. In this chapter, I introduce the concept of moral behavior for an automated vehicle, argue the need for research in this area through responses to anticipated critiques, and discuss relevant applications from machine ethics and moral modeling research. 1 Ethical Decision Making for Automated Vehicles Vehicle automation has progressed rapidly this millennium, mirroring improvements in machine learning, sensing, and processing. Media coverage often focuses on the anticipated safety benefits from automation, as computers are expected to be more attentive, precise, and predictable than human drivers. Mentioned less often are the novel problems from automated vehicle crash. The first problem is liability, as it is currently unclear who would be at fault if a vehicle crashed while self-driving. The second problem is the ability of an automated vehicle to make ethically-complex decisions when driving, particularly prior to a crash. This chapter focuses on the second problem, and the application of machine ethics to vehicle automation. Driving at any significant speed can never be completely safe. A loaded tractor trailer at 100 km/hr requires eight seconds to come to a complete stop, and a passenger car requires three seconds [1]. Truly safe travel requires accurate predictions of other vehicle behavior over this time frame, something that is simply not possible given the close proximities of road vehicles. To ensure its own safety, an automated vehicle must continually assess risk: the risk of traveling a certain speed on a certain curve, of crossing the centerline to