Gaussian path model library for intuitive robot motion programming by demonstration

2025-09-12

AI Summary

To address the limited capability in learning and reproducing complex trajectories in robot motion programming, this paper proposes the Gaussian Path Library (GPL) framework. GPL leverages human demonstration data and integrates Gaussian process regression, geometric path shape feature extraction, and pattern classification to automatically identify and model diverse geometric trajectory classes. Furthermore, it incorporates a geometric analysis mechanism enabling demonstration-driven online editing and adaptive refinement of existing path models. Experimental results demonstrate that GPL achieves high-accuracy classification (>96%) and smooth, robust trajectory reproduction across various nonlinear paths—including spirals, loops, and piecewise curves. The framework significantly enhances the intuitiveness, flexibility, and scalability of programming-by-demonstration for robotic manipulation tasks.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Machine Learning: Learning with Manifolds

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

This paper presents a system for generating Gaussian path models from teaching data representing the path shape. In addition, methods for using these path models to classify human demonstrations of paths are introduced. By generating a library of multiple Gaussian path models of various shapes, human demonstrations can be used for intuitive robot motion programming. A method for modifying existing Gaussian path models by demonstration through geometric analysis is also presented.

Problem

• Generating Gaussian path models from teaching data
• Classifying human demonstrations using path models
• Modifying existing path models through geometric analysis

Innovation

• Gaussian path models from teaching data
• Classify human demonstrations for paths
• Modify path models via geometric analysis