no code implementations • 27 Jul 2022 • Xiang Zhu, Shucheng Kang, Jianyu Chen
In this paper, we propose a contact-safe reinforcement learning framework for contact-rich robot manipulation, which maintains safety in both the task space and joint space.