Jacta: A Versatile Planner for Learning Dexterous and Whole-body Manipulation

TL;DR

Jacta is a versatile motion planner that generates demonstrations to facilitate reinforcement learning in complex dexterous and whole-body robotic manipulation tasks, enabling transfer to real robots.

Contribution

It introduces a flexible, system-agnostic planner that simplifies data collection and improves learning efficiency for complex manipulation tasks.

Findings

Efficient policy learning for complex manipulation

Demonstrations transfer effectively to real robots

Outperforms traditional reinforcement learning methods

Abstract

Robotic manipulation is challenging due to discontinuous dynamics, as well as high-dimensional state and action spaces. Data-driven approaches that succeed in manipulation tasks require large amounts of data and expert demonstrations, typically from humans. Existing planners are restricted to specific systems and often depend on specialized algorithms for using demonstrations. Therefore, we introduce a flexible motion planner tailored to dexterous and whole-body manipulation tasks. Our planner creates readily usable demonstrations for reinforcement learning algorithms, eliminating the need for additional training pipeline complexities. With this approach, we can efficiently learn policies for complex manipulation tasks, where traditional reinforcement learning alone only makes little progress. Furthermore, we demonstrate that learned policies are transferable to real robotic systems for solving complex dexterous manipulation tasks. Project website: https://jacta-manipulation.github.io/