Single-Shot Learning of Stable Dynamical Systems for Long-Horizon Manipulation Tasks

TL;DR

This paper presents a new method for learning stable dynamical systems for long-horizon robotic manipulation tasks, improving success rates and reducing training data requirements by segmenting tasks and ensuring global stability.

Contribution

It introduces a novel approach that segments demonstrations into waypoints and learns globally stable policies for each segment, enhancing robustness and efficiency in long-horizon tasks.

Findings

Effective transfer from simulation to real robots

Improved task success rates with less training data

Robust performance under noise and disturbances

Abstract

Mastering complex sequential tasks continues to pose a significant challenge in robotics. While there has been progress in learning long-horizon manipulation tasks, most existing approaches lack rigorous mathematical guarantees for ensuring reliable and successful execution. In this paper, we extend previous work on learning long-horizon tasks and stable policies, focusing on improving task success rates while reducing the amount of training data needed. Our approach introduces a novel method that (1) segments long-horizon demonstrations into discrete steps defined by waypoints and subgoals, and (2) learns globally stable dynamical system policies to guide the robot to each subgoal, even in the face of sensory noise and random disturbances. We validate our approach through both simulation and real-world experiments, demonstrating effective transfer from simulation to physical robotic platforms. Code is available at https://github.com/Alestaubin/stable-imitation-policy-with-waypoints