Augmenting Reinforcement Learning with Behavior Primitives for Diverse Manipulation Tasks

TL;DR

This paper introduces MAPLE, a framework that enhances reinforcement learning for manipulation tasks by integrating a library of behavior primitives, leading to improved performance, transferability, and interpretability.

Contribution

The paper proposes a hierarchical reinforcement learning method augmented with behavior primitives, enabling better long-horizon manipulation and transfer to new tasks and hardware.

Findings

MAPLE significantly outperforms baseline methods in simulated tasks.

The learned behaviors exhibit a clear compositional structure.

Policies transfer effectively to new task variants and real hardware.

Abstract

Realistic manipulation tasks require a robot to interact with an environment with a prolonged sequence of motor actions. While deep reinforcement learning methods have recently emerged as a promising paradigm for automating manipulation behaviors, they usually fall short in long-horizon tasks due to the exploration burden. This work introduces Manipulation Primitive-augmented reinforcement Learning (MAPLE), a learning framework that augments standard reinforcement learning algorithms with a pre-defined library of behavior primitives. These behavior primitives are robust functional modules specialized in achieving manipulation goals, such as grasping and pushing. To use these heterogeneous primitives, we develop a hierarchical policy that involves the primitives and instantiates their executions with input parameters. We demonstrate that MAPLE outperforms baseline approaches by a significant margin on a suite of simulated manipulation tasks. We also quantify the compositional structure of the learned behaviors and highlight our method's ability to transfer policies to new task variants and to physical hardware. Videos and code are available at https://ut-austin-rpl.github.io/maple