Learning Diffusion Policies for Robotic Manipulation of Timber Joinery under Fabrication Uncertainty

2025-11-21

AI Summary

High contact-assembly uncertainty in timber mortise-and-tenon joints arises from fabrication errors (e.g., ±10 mm tenon positional deviations) and material defects, hindering reliable robotic construction in real-world settings. Method: This paper proposes a perception–motor control method based on diffusion policy learning. We introduce a sensory–motor diffusion policy framework specifically designed for large-scale, high-contact assembly tasks, and—uniquely in simulation—explicitly model stochastic fabrication disturbances to ensure strong generalization and robustness. Contribution/Results: Experiments demonstrate 100% success rate under nominal conditions and a mean overall success rate of 75%, significantly outperforming conventional approaches. Our method overcomes the adaptability bottleneck of robotic construction in uncertain, real-world fabrication environments. It establishes a scalable, perception–action co-design paradigm for intelligent timber construction, enabling robust physical interaction despite geometric and material variability.

Technology Category

• Intelligent Robots: Manipulation
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Construction uncertainties such as fabrication inaccuracies and material imperfections pose a significant challenge to contact-rich robotic manipulation by hindering precise and robust assembly. In this paper, we explore the performance and robustness of diffusion policy learning as a promising solution for contact-sensitive robotic assembly at construction scale, using timber mortise and tenon joints as a case study. A two-phase study is conducted: first, to evaluate policy performance and applicability; second, to assess robustness in handling fabrication uncertainties simulated as randomized perturbations to the mortise position. The best-performing policy achieved a total average success rate of 75% with perturbations up to 10 mm, including 100% success in unperturbed cases. The results demonstrate the potential of sensory-motor diffusion policies to generalize to a wide range of complex, contact-rich assembly tasks across construction and manufacturing, advancing robotic construction under uncertainty and contributing to safer, more efficient building practices.

Problem

• Robotic manipulation under fabrication uncertainties
• Diffusion policies for contact-rich timber joinery assembly
• Handling positional perturbations in construction-scale assembly tasks

Innovation

• Learning diffusion policies for robotic manipulation
• Handling fabrication uncertainties with sensory-motor policies
• Achieving robust assembly through contact-sensitive diffusion learning