Collision-free Path Planning in the Latent Space through cGANs

TL;DR

This paper introduces a novel collision-free path planning method using conditional Generative Adversarial Networks (cGANs) to map the latent space to collision-free regions, enabling flexible and efficient path generation for robotic arms.

Contribution

The paper presents a new approach that leverages cGANs to directly model collision-free regions in the robot's joint space, facilitating real-time path planning with any optimizer.

Findings

Successfully verified on a simulated two-link robot arm

Demonstrated effective mapping of collision-free space

Enabled flexible path planning with various optimization criteria

Abstract

We show a new method for collision-free path planning by cGANs by mapping its latent space to only the collision-free areas of the robot joint space. Our method simply provides this collision-free latent space after which any planner, using any optimization conditions, can be used to generate the most suitable paths on the fly. We successfully verified this method with a simulated two-link robot arm.