Learning Robotic Manipulation of Granular Media

TL;DR

This paper investigates robotic manipulation of granular media, evaluating predictive models for scooping and dumping, and introduces a tailored convolutional network that improves physical interaction modeling and manipulation accuracy.

Contribution

The paper presents a novel convolutional network architecture optimized for granular media manipulation, demonstrating superior predictive accuracy and control performance.

Findings

The tailored convolutional model accurately predicts media dynamics.

Explicit physics prediction outperforms hand-crafted and implicit models.

The approach enables more precise and effective manipulation of granular media.

Abstract

In this paper, we examine the problem of robotic manipulation of granular media. We evaluate multiple predictive models used to infer the dynamics of scooping and dumping actions. These models are evaluated on a task that involves manipulating the media in order to deform it into a desired shape. Our best performing model is based on a highly-tailored convolutional network architecture with domain-specific optimizations, which we show accurately models the physical interaction of the robotic scoop with the underlying media. We empirically demonstrate that explicitly predicting physical mechanics results in a policy that out-performs both a hand-crafted dynamics baseline, and a "value-network", which must otherwise implicitly predict the same mechanics in order to produce accurate value estimates.