ForceMimic: Force-Centric Imitation Learning with Force-Motion Capture System for Contact-Rich Manipulation

TL;DR

ForceMimic introduces a force-centric imitation learning system that enhances contact-rich manipulation tasks by leveraging force-aware demonstrations, resulting in more robust policies and higher success rates compared to vision-only methods.

Contribution

The paper presents a novel force-centric imitation learning framework with a dedicated demonstration collection system and a hybrid force-motion algorithm for improved contact-rich manipulation.

Findings

Achieved a 54.5% higher success rate in vegetable peeling tasks.

Demonstrated faster data collection with the ForceCapture system (5 minutes).

Enabled robust manipulation policies through hybrid force-position control.

Abstract

In most contact-rich manipulation tasks, humans apply time-varying forces to the target object, compensating for inaccuracies in the vision-guided hand trajectory. However, current robot learning algorithms primarily focus on trajectory-based policy, with limited attention given to learning force-related skills. To address this limitation, we introduce ForceMimic, a force-centric robot learning system, providing a natural, force-aware and robot-free robotic demonstration collection system, along with a hybrid force-motion imitation learning algorithm for robust contact-rich manipulation. Using the proposed ForceCapture system, an operator can peel a zucchini in 5 minutes, while force-feedback teleoperation takes over 13 minutes and struggles with task completion. With the collected data, we propose HybridIL to train a force-centric imitation learning model, equipped with hybrid force-position control primitive to fit the predicted wrench-position parameters during robot execution. Experiments demonstrate that our approach enables the model to learn a more robust policy under the contact-rich task of vegetable peeling, increasing the success rates by 54.5% relatively compared to state-ofthe-art pure-vision-based imitation learning. Hardware, code, data and more results can be found on the project website at https://forcemimic.github.io.