Offline Learning of Counterfactual Predictions for Real-World Robotic Reinforcement Learning

TL;DR

This paper introduces an offline learning approach for robotic manipulation that uses counterfactual predictions from visual data to improve online reinforcement learning efficiency, especially in contact-rich tasks.

Contribution

It proposes a novel method combining offline learned counterfactual predictions with force feedback to enhance sample efficiency in real-world robotic RL.

Findings

Counterfactual predictions improve sample efficiency.

Combining visual and force data enhances policy learning.

Method performs well in simulation and real-world experiments.

Abstract

We consider real-world reinforcement learning (RL) of robotic manipulation tasks that involve both visuomotor skills and contact-rich skills. We aim to train a policy that maps multimodal sensory observations (vision and force) to a manipulator's joint velocities under practical considerations. We propose to use offline samples to learn a set of general value functions (GVFs) that make counterfactual predictions from the visual inputs. We show that combining the offline learned counterfactual predictions with force feedbacks in online policy learning allows efficient reinforcement learning given only a terminal (success/failure) reward. We argue that the learned counterfactual predictions form a compact and informative representation that enables sample efficiency and provides auxiliary reward signals that guide online explorations towards contact-rich states. Various experiments in simulation and real-world settings were performed for evaluation. Recordings of the real-world robot training can be found via https://sites.google.com/view/realrl.