Communicating Missing Causal Information to Explain a Robot’s Past Behavior

Robots need to explain their behavior gain trust. Existing research has focused on explaining a robot’s current behavior, yet it remains unknown challenging how provide explanations of past actions in an environment that might change after actions, leading critical missing causal information due moved objects. We conducted experiment (N = 665) investigating robot could help participants infer the by replaying physically, using verbal explanations, and projecting visual onto environment. Participants watched videos its completion integrated mobile kitting task. During replay, objects are already gone, so needed where object was picked, ground obstacle had been, placed. Based results, we recommend combining physical replay with speech projection indicators (Replay-Project-Say) all (picking, navigation, placement) from actions. This condition best outcome both task-based—effectiveness, efficiency, confidence—and team-based metrics—workload If one’s focus is then markers for navigation inferences placing inferences.