Robotic Table Wiping via Reinforcement Learning and Whole-body Trajectory Optimization

TL;DR

This paper presents a reinforcement learning and trajectory optimization framework enabling assistive robots to autonomously wipe tables, effectively handling uncertain spill dynamics and ensuring safe operation in cluttered environments.

Contribution

It introduces a novel stochastic model for spill dynamics, combines RL with trajectory optimization for zero-shot sim-to-real transfer, and validates the approach on real robots.

Findings

Successful zero-shot transfer from simulation to real robot wiping tasks

Effective handling of uncertain spill dynamics in cluttered environments

Robust wiping performance demonstrated in hardware experiments

Abstract

We propose a framework to enable multipurpose assistive mobile robots to autonomously wipe tables to clean spills and crumbs. This problem is challenging, as it requires planning wiping actions while reasoning over uncertain latent dynamics of crumbs and spills captured via high-dimensional visual observations. Simultaneously, we must guarantee constraints satisfaction to enable safe deployment in unstructured cluttered environments. To tackle this problem, we first propose a stochastic differential equation to model crumbs and spill dynamics and absorption with a robot wiper. Using this model, we train a vision-based policy for planning wiping actions in simulation using reinforcement learning (RL). To enable zero-shot sim-to-real deployment, we dovetail the RL policy with a whole-body trajectory optimization framework to compute base and arm joint trajectories that execute the desired wiping motions while guaranteeing constraints satisfaction. We extensively validate our approach in simulation and on hardware. Video: https://youtu.be/inORKP4F3EI