ArrayBot: Reinforcement Learning for Generalizable Distributed Manipulation through Touch
This work addresses the challenge of scalable and adaptable robotic manipulation for tabletop tasks, representing a novel method for handling redundant actions in distributed systems.
The researchers tackled the problem of generalizable distributed manipulation using a tactile sensor array, and achieved a system that can relocate diverse objects with policies that generalize to unseen shapes and transfer to a physical robot without domain randomization.
We present ArrayBot, a distributed manipulation system consisting of a $16 \times 16$ array of vertically sliding pillars integrated with tactile sensors, which can simultaneously support, perceive, and manipulate the tabletop objects. Towards generalizable distributed manipulation, we leverage reinforcement learning (RL) algorithms for the automatic discovery of control policies. In the face of the massively redundant actions, we propose to reshape the action space by considering the spatially local action patch and the low-frequency actions in the frequency domain. With this reshaped action space, we train RL agents that can relocate diverse objects through tactile observations only. Surprisingly, we find that the discovered policy can not only generalize to unseen object shapes in the simulator but also transfer to the physical robot without any domain randomization. Leveraging the deployed policy, we present abundant real-world manipulation tasks, illustrating the vast potential of RL on ArrayBot for distributed manipulation.