Sim-to-Real gap in RL: Use Case with TIAGo and Isaac Sim/Gym

TL;DR

This paper investigates the sim-to-real transfer of reinforcement learning policies for robotic manipulation using TIAGo, comparing Nvidia's Isaac Gym and Isaac Sim simulators, and demonstrates successful transfer with collision-free movements.

Contribution

It provides a comparative analysis of two advanced simulators for RL-based robotic manipulation and demonstrates effective sim-to-real transfer for collision-free control.

Findings

Successful transfer of RL policies from simulation to real robot

Comparable collision-free movements in simulation and real environment

Insights into control architectures for sim-to-real transfer

Abstract

This paper explores policy-learning approaches in the context of sim-to-real transfer for robotic manipulation using a TIAGo mobile manipulator, focusing on two state-of-art simulators, Isaac Gym and Isaac Sim, both developed by Nvidia. Control architectures are discussed, with a particular emphasis on achieving collision-less movement in both simulation and the real environment. Presented results demonstrate successful sim-to-real transfer, showcasing similar movements executed by an RL-trained model in both simulated and real setups.