A Framework for Learning and Reusing Robotic Skills

TL;DR

This paper introduces a framework for creating a library of robotic motion primitives that can be learned from demonstration, segmented, clustered, and reused in new environments through trajectory editing, demonstrated via simulation.

Contribution

It proposes novel multimodal segmentation and trajectory clustering methods for robotic skill learning and reuse, enhancing adaptability and efficiency.

Findings

Successful simulation results demonstrating the framework

Effective segmentation and clustering of demonstrated skills

Trajectory editing enables skill transfer to new environments

Abstract

In this paper, we present our work in progress towards creating a library of motion primitives. This library facilitates easier and more intuitive learning and reusing of robotic skills. Users can teach robots complex skills through Learning from Demonstration, which is automatically segmented into primitives and stored in clusters of similar skills. We propose a novel multimodal segmentation method as well as a novel trajectory clustering method. Then, when needed for reuse, we transform primitives into new environments using trajectory editing. We present simulated results for our framework with demonstrations taken on real-world robots.