Liability, Ethics, and Culture-Aware Behavior Specification using Rulebooks
This addresses the problem of behavior specification for autonomous systems, particularly self-driving cars, by providing a structured way to handle ambiguous and competing goals, though it appears incremental as it builds on existing rule-based approaches.
The paper tackles the challenge of defining behavior for autonomous agents like self-driving cars that must satisfy conflicting objectives from law, ethics, and culture, by introducing a 'rulebook' method that imposes a pre-order on outcomes based on prioritized rules, and demonstrates its application in self-driving domains while showing domain independence.
The behavior of self-driving cars must be compatible with an enormous set of conflicting and ambiguous objectives, from law, from ethics, from the local culture, and so on. This paper describes a new way to conveniently define the desired behavior for autonomous agents, which we use on the self-driving cars developed at nuTonomy. We define a "rulebook" as a pre-ordered set of "rules", each akin to a violation metric on the possible outcomes ("realizations"). The rules are partially ordered by priority. The semantics of a rulebook imposes a pre-order on the set of realizations. We study the compositional properties of the rulebooks, and we derive which operations we can allow on the rulebooks to preserve previously-introduced constraints. While we demonstrate the application of these techniques in the self-driving domain, the methods are domain-independent.