WireCraft: A Simulation Benchmark for Industrial DLO Manipulation

2026-06-16

AI Summary

This work addresses the lack of a unified, reproducible, and realistic simulation benchmark for deformable linear object (DLO) manipulation in industrial settings, which has hindered policy development and fair evaluation. To bridge this gap, the authors introduce a modular simulation benchmark encompassing three industrial-grade tasks: connector insertion, clip routing, and channel alignment. The platform supports configurable difficulty levels and asset substitution, and uniquely integrates both articulated and deformable physical models, multimodal policy frameworks—including reinforcement learning, imitation learning, and vision-language-action approaches—and a simulation-to-reality alignment mechanism. Experiments demonstrate that reinforcement learning policies leveraging privileged state information achieve success rates exceeding 82% across all tasks, whereas purely vision-based methods still struggle with fine alignment, revealing current limitations in vision-driven strategies. This benchmark provides an open and standardized foundation for DLO manipulation research.

Abstract

Deformable Linear Objects (DLOs), such as wires and cables, are central to industrial assembly. Unlike rigid objects, whose state is captured by a 6-DoF pose, DLOs have an infinite-dimensional configuration space and deform continuously under contact with grippers, fixtures, and the workspace, making them a demanding benchmark for general dexterous manipulation. Despite their importance, policy development and comparison remain difficult: existing benchmarks are often tied to specific hardware setups, lack modular and customizable task assets, or study generic deformable-object tasks without the fixtures relevant to real-world industrial wire manipulation. Few benchmarks align simulation, real-world data, and shared evaluation protocols. To bridge this gap, we introduce WireCraft, a simulation benchmark for industrial DLO manipulation with configurable difficulty and assets, spanning three task families: connector insertion, clip routing, and channel seating. It supports two complementary DLO physics models, articulated and deformable, and the trajectories come from both simulation and a physical UR5. We benchmark reinforcement learning (RL), imitation learning (IL), and vision-language-action (VLA) policies under shared metrics. Privileged state-based RL solves a representative setting in each task family with over 82\% success, confirming the tasks are well-posed. For connector insertion, however, the transition from reaching the socket to contact-rich alignment remains a key bottleneck for vision RL, IL, and VLA policies. These results indicate that industrial DLO manipulation, though tractable under privileged state, remains an open challenge for current vision-based learning. The benchmark, data, and tools will be open-sourced upon acceptance.

Problem

• Deformable Linear Objects
• Industrial Manipulation
• Simulation Benchmark
• Dexterous Manipulation
• Vision-based Learning

Innovation

• Deformable Linear Objects
• Simulation Benchmark
• Industrial Manipulation
• Vision-Based Learning
• Physics Modeling