I-ORCA – Implicitly Nudging Human Trajectories
Date:
Presented work done during the summer internship at Honda Research Institute in San Jose, CA. The work tackled a novel problem: How should we design robot trajectories that implicitly nudge human trajectories in desired directions using their collision avoidance behavior. Toward this, we inverted the well-known Optimal Reciprocal Collision Avoidance (ORCA) algorithm for modeling collision-avoiding agents.