High Precision Real Time Collision Detection

TL;DR

This paper introduces a high-precision, real-time collision detection system for robotics that requires only URDF and extends the GJK algorithm, demonstrating superior performance in self-collision scenarios.

Contribution

A novel collision detection system that is generic, highly precise, real-time capable, and easy to deploy using only URDF files, with an improved GJK algorithm extension.

Findings

Successfully detects self-collisions in KUKA LBR iiwa and Mecademic Meca500.

Outperforms the Flexible Collision Library (FCL) in performance.

Operates efficiently with minimal setup requirements.

Abstract

Collision detection and collision avoidance are essential components in these systems for safe human-robot interactions. Robotics systems that can work "out-of-the-box" without excessive amount of installation and calibration from the experts is highly ideal. For this, we propose a generic, high precision, collision detect system that only requires the unified robot description format (URDF) and is capable of running in real time. We extended the Gilbert-Johnson-Keerthi (GJK) algorithm by utilizing a geometrical approach to determine the distance between each rigid body in the environment and check for collisions. The proposed system's performance is shown by checking the self-collision of the KUKA LBR iiwa 7 R800 and the Mecademic Meca500. The performance is compared to the Flexible Collision Library (FCL).