Simulation of Optical Tactile Sensors Supporting Slip and Rotation using Path Tracing and IMPM

TL;DR

This paper presents a novel simulation approach for optical tactile sensors that combines path tracing and IMPM to accurately model slip and rotation, enhancing robotic manipulation capabilities.

Contribution

It introduces a new simulation method integrating path tracing and IMPM for realistic optical tactile sensor modeling, including slip and rotation effects.

Findings

Path tracing improves similarity to real tactile data.

IMPM enables accurate simulation of slip and rotation.

The method enhances robotic manipulation simulation fidelity.

Abstract

Optical tactile sensors are extensively utilized in intelligent robot manipulation due to their ability to acquire high-resolution tactile information at a lower cost. However, achieving adequate reality and versatility in simulating optical tactile sensors is challenging. In this paper, we propose a simulation method and validate its effectiveness through experiments. We utilize path tracing for image rendering, achieving higher similarity to real data than the baseline method in simulating pressing scenarios. Additionally, we apply the improved Material Point Method(IMPM) algorithm to simulate the relative rest between the object and the elastomer surface when the object is in motion, enabling more accurate simulation of complex manipulations such as slip and rotation.