AI-based Framework for Robust Model-Based Connector Mating in Robotic Wire Harness Installation

2025-03-12

AI Summary

Automotive wiring harness connector insertion—critical for vehicle assembly—has long relied on manual labor due to stringent requirements for sub-millimeter precision and robust environmental adaptability, hindering automation. Method: This paper proposes a robotized insertion method integrating real-time force control with deep visuo-tactile learning. We introduce the first multimodal Transformer architecture unifying visual, tactile, and proprioceptive sensory streams; design an expert-free pipeline for autonomous data collection and policy optimization; and generate auditable, certifiable native industrial controller code. Results: Evaluated on center-console assembly tasks, our approach reduces cycle time significantly while achieving higher insertion success rates and superior environmental robustness compared to conventional teach-pendant programming and offline trajectory planning. It delivers a production-ready intelligent manipulation paradigm for flexible automotive assembly lines.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling
• Computer Vision: Vision for Robotics & Autonomous Driving

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Despite the widespread adoption of industrial robots in automotive assembly, wire harness installation remains a largely manual process, as it requires precise and flexible manipulation. To address this challenge, we design a novel AI-based framework that automates cable connector mating by integrating force control with deep visuotactile learning. Our system optimizes search-and-insertion strategies using first-order optimization over a multimodal transformer architecture trained on visual, tactile, and proprioceptive data. Additionally, we design a novel automated data collection and optimization pipeline that minimizes the need for machine learning expertise. The framework optimizes robot programs that run natively on standard industrial controllers, permitting human experts to audit and certify them. Experimental validations on a center console assembly task demonstrate significant improvements in cycle times and robustness compared to conventional robot programming approaches. Videos are available under https://claudius-kienle.github.io/AppMuTT.

Problem

• Automates cable connector mating in robotic wire harness installation
• Integrates force control with deep visuotactile learning for precise manipulation
• Optimizes robot programs for industrial controllers, improving cycle times and robustness

Innovation

• AI-based framework automates cable connector mating
• Integrates force control with deep visuotactile learning
• Optimizes robot programs for industrial controllers