Human Inspired Effort Distribution During Collision Avoidance in Human-Robot Motion

Abstract

Recent works in the area of human robot motion showed that behaving in a human-like manner allows a robot to reduce global cognitive effort for people in the environment. Given that collision avoidance situations between people are solved cooperatively, this work models the manner in which this cooperation is done so that a robot can replicate their behavior. To that end, hundreds of situations where two walkers have crossing trajectories were analyzed. Based on these human trajectories involving a collision avoidance task, we determined how total effort is shared between each walker depending on several factors of the interaction such as crossing angle, time to collision and speed. To validate our approach, a proof of concept is integrated into ROS with Reciprocal Velocity Objects (RVO) in order to distribute collision avoidance effort in a human-like way.

Effective Human-Robot Collaboration in near symmetry collision scenarios

“…In our collision avoidance dataset, described in details in [16], we evaluate dyadic interactions between people when avoiding collisions with each other. The experiments were composed of over 450 runs divided about equally into five distinct crossing angles: 30°, 60°, 90°, 120°and 150°.…”