Applying Ethical Decision Making on Space Missions

As the applications for autonomous systems in space missions grows, there is a great deal of interest in assuring ethical behaviour. If humans are to cooperate with autonomous systems, there must be assurance that they are trustworthy in making ethical decisions. In particular, this applies to situations in which there is uncertainty over the outcomes of decisions. Our poster presents an explainable, expressive framework designed for this type of decision making. Based on Sven-Ove Hansson's Hypothetical Retrospection, we present a procedure which fairly evaluates actions comparatively though morally relevant information. We demonstrate the procedure in a Mars mission scenario, in which a rover chooses whether to investigate the status of a crew, or continue acting as a communication relay. The procedure selects different explainable actions dependent on a predetermined configuration of moral principles.