The Ingredients of Real-World Robotic Reinforcement Learning

TL;DR

This paper presents a scalable, autonomous reinforcement learning system for real-world robotic manipulation that learns without human intervention, using only onboard perception and simple reward signals.

Contribution

It introduces a novel system capable of continuous autonomous learning in real-world robotics, addressing challenges like no manual resets or engineered rewards.

Findings

Successfully learned vision-based manipulation skills with a three-fingered hand

Demonstrated continuous learning without human intervention

Provided an in-depth analysis of real-world RL challenges

Abstract

The success of reinforcement learning for real world robotics has been, in many cases limited to instrumented laboratory scenarios, often requiring arduous human effort and oversight to enable continuous learning. In this work, we discuss the elements that are needed for a robotic learning system that can continually and autonomously improve with data collected in the real world. We propose a particular instantiation of such a system, using dexterous manipulation as our case study. Subsequently, we investigate a number of challenges that come up when learning without instrumentation. In such settings, learning must be feasible without manually designed resets, using only on-board perception, and without hand-engineered reward functions. We propose simple and scalable solutions to these challenges, and then demonstrate the efficacy of our proposed system on a set of dexterous robotic manipulation tasks, providing an in-depth analysis of the challenges associated with this learning paradigm. We demonstrate that our complete system can learn without any human intervention, acquiring a variety of vision-based skills with a real-world three-fingered hand. Results and videos can be found at https://sites.google.com/view/realworld-rl/