Imitation Learning-Based Path Generation for the Complex Assembly of Deformable Objects

2025-05-30

AI Summary

High-precision motion planning for deformable object assembly typically relies on complex, computationally expensive dynamical modeling. Method: This paper proposes a lightweight motion planning framework integrating offline path initialization, human-in-the-loop correction, and behavior cloning (BC). It employs a simplified deformable-body dynamics model coupled with human demonstration-driven imitation learning, operating under low-dimensional state representation and compliant control. The pipeline comprises collision-free offline path generation, compliant robotic execution, BC-based policy learning from expert demonstrations, and human-robot collaborative data augmentation. Contribution/Results: Evaluated across diverse soft-object assembly tasks, the method achieves a 42% improvement in assembly success rate and an 83% reduction in planning latency, while exhibiting strong policy generalization. By avoiding full-scale dynamical modeling, it significantly reduces both modeling complexity and real-time computational overhead.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

This paper investigates how learning can be used to ease the design of high-quality paths for the assembly of deformable objects. Object dynamics plays an important role when manipulating deformable objects; thus, detailed models are often used when conducting motion planning for deformable objects. We propose to use human demonstrations and learning to enable motion planning of deformable objects with only simple dynamical models of the objects. In particular, we use the offline collision-free path planning, to generate a large number of reference paths based on a simple model of the deformable object. Subsequently, we execute the collision-free paths on a robot with a compliant control such that a human can slightly modify the path to complete the task successfully. Finally, based on the virtual path data sets and the human corrected ones, we use behavior cloning (BC) to create a dexterous policy that follows one reference path to finish a given task.

Problem

• Learning-based path generation for deformable object assembly
• Reducing reliance on complex dynamical models via human demonstrations
• Behavior cloning to create policies from human-corrected paths

Innovation

• Uses imitation learning for deformable objects
• Combines simple models with human demonstrations
• Applies behavior cloning for path correction