An Incremental Inverse Reinforcement Learning Approach for Motion Planning with Human Preferences

Humans often demonstrate diverse behaviors due to their personal preferences, for instance related to their individual execution style or personal margin for safety. In this paper, we consider the problem of integrating such preferences into trajectory planning for robotic manipulators. We first learn reward functions that represent the user path and motion preferences from kinesthetic demonstration. We then use a discrete-time trajectory optimization scheme to produce trajectories that adhere to both task requirements and user preferences. We go beyond the state of art by designing a feature set that captures the fundamental preferences in a manipulation task, such as timing of the motion. We further demonstrate that our method is capable of generalizing such preferences to new scenarios. We implement our algorithm on a Franka Emika 7-DoF robot arm, and validate the functionality and flexibility of our approach in a user study. The results show that non-expert users are able to teach the robot their preferences with just a few iterations of feedback.