The Ingredients of Real-World Robotic Reinforcement Learning

04/27/2020
by   Henry Zhu, et al.
14

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/

READ FULL TEXT

page 3

page 8

page 10

page 16

page 17

page 18

page 19

page 20

research
10/07/2019

DexPilot: Vision Based Teleoperation of Dexterous Robotic Hand-Arm System

Teleoperation offers the possibility of imparting robotic systems with s...
research
03/29/2022

Demonstration-Bootstrapped Autonomous Practicing via Multi-Task Reinforcement Learning

Reinforcement learning systems have the potential to enable continuous i...
research
07/07/2023

Polybot: Training One Policy Across Robots While Embracing Variability

Reusing large datasets is crucial to scale vision-based robotic manipula...
research
12/13/2022

Learning Robotic Navigation from Experience: Principles, Methods, and Recent Results

Navigation is one of the most heavily studied problems in robotics, and ...
research
03/15/2022

Vision-Based Manipulators Need to Also See from Their Hands

We study how the choice of visual perspective affects learning and gener...
research
01/04/2021

Machine Learning for Robotic Manipulation

The past decade has witnessed the tremendous successes of machine learni...
research
07/03/2023

MoVie: Visual Model-Based Policy Adaptation for View Generalization

Visual Reinforcement Learning (RL) agents trained on limited views face ...

Please sign up or login with your details

Forgot password? Click here to reset