MatchMaker: Automated Asset Generation for Robotic Assembly

2025-03-07

AI Summary

Robot assembly simulation suffers from labor-intensive, low-diversity asset construction, severely hindering policy learning for multi-task generalist agents. To address this, we propose the first fully automated framework for generating paired assembly assets, featuring three core capabilities: (1) repair of geometrically incompatible CAD models, (2) single-part pairing generation, and (3) gap-controllable surface etching. We introduce a novel geometry-aware pairing generation method tailored to functional assembly requirements and a tunable-gap modeling approach, integrating computational geometry analysis, implicit surface modeling, contact-surface topology optimization, and parametric gap synthesis—all end-to-end driven by physics-based simulation constraints. Experiments demonstrate that our generated asset pairs significantly outperform existing methods in structural diversity and downstream assembly policy performance, markedly improving sim-to-real generalization.

Technology Category

• Intelligent Robots: Manipulation
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Robotic assembly remains a significant challenge due to complexities in visual perception, functional grasping, contact-rich manipulation, and performing high-precision tasks. Simulation-based learning and sim-to-real transfer have led to recent success in solving assembly tasks in the presence of object pose variation, perception noise, and control error; however, the development of a generalist (i.e., multi-task) agent for a broad range of assembly tasks has been limited by the need to manually curate assembly assets, which greatly constrains the number and diversity of assembly problems that can be used for policy learning. Inspired by recent success of using generative AI to scale up robot learning, we propose MatchMaker, a pipeline to automatically generate diverse, simulation-compatible assembly asset pairs to facilitate learning assembly skills. Specifically, MatchMaker can 1) take a simulation-incompatible, interpenetrating asset pair as input, and automatically convert it into a simulation-compatible, interpenetration-free pair, 2) take an arbitrary single asset as input, and generate a geometrically-mating asset to create an asset pair, 3) automatically erode contact surfaces from (1) or (2) according to a user-specified clearance parameter to generate realistic parts. We demonstrate that data generated by MatchMaker outperforms previous work in terms of diversity and effectiveness for downstream assembly skill learning. For videos and additional details, please see our project website: https://wangyian-me.github.io/MatchMaker/.

Problem

• Automates generation of diverse assembly assets for robotic learning.
• Converts incompatible assets into simulation-compatible, interpenetration-free pairs.
• Enhances assembly skill learning with realistic, geometrically-mating asset pairs.

Innovation

• Automated generation of simulation-compatible assembly assets
• Generative AI for diverse assembly asset creation
• Automatic erosion of contact surfaces for realistic parts