A Framework for Robot Manipulation: Skill Formalism, Meta Learning and Adaptive Control

TL;DR

This paper presents a new framework for robot manipulation that combines skill formalism, meta learning, and adaptive control, enabling rapid learning and execution of complex, force-sensitive tasks with minimal training time.

Contribution

It introduces a formalism for expressing and learning manipulation skills that incorporates expert knowledge and quality metrics, improving task complexity reduction and learning efficiency.

Findings

System learns peg-in-hole tasks in under 20 minutes.

Achieves faster task execution than humans in experiments.

Significantly reduces manipulation complexity with the proposed framework.

Abstract

In this paper we introduce a novel framework for expressing and learning force-sensitive robot manipulation skills. It is based on a formalism that extends our previous work on adaptive impedance control with meta parameter learning and compatible skill specifications. This way the system is also able to make use of abstract expert knowledge by incorporating process descriptions and quality evaluation metrics. We evaluate various state-of-the-art schemes for the meta parameter learning and experimentally compare selected ones. Our results clearly indicate that the combination of our adaptive impedance controller with a carefully defined skill formalism significantly reduces the complexity of manipulation tasks even for learning peg-in-hole with submillimeter industrial tolerances. Overall, the considered system is able to learn variations of this skill in under 20 minutes. In fact, experimentally the system was able to perform the learned tasks faster than humans, leading to the first learning-based solution of complex assembly at such real-world performance.