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

Claudius Kienle1,2,†, Benjamin Alt1,3,†, Finn Schneider1,4, Tobias Pertlwieser4, Rainer Jäkel1, Rania Rayyes4,
1ArtiMinds Robotics, Karlsruhe, Germany 2IAS Lab, Computer Science Department, TU Darmstadt, Germany 3AICOR Institute for Artificial Intelligence, University of Bremen, Germany 4Institute for Material Handling and Logistics (IFL), Karlsruhe Institute of Technology (KIT), Karlsruhe, Germany
Paper arXiv Video

Robotic mating of electrical connectors in a center console: Using our learned, neural model of the search-and-insertion strategy, the search pattern (bottom left, red) is optimized (green) to ensure robust installation with minimal temporal overhead.

Abstract

Despite the widespread adoption of industrial robots in automotive assembly, wire harness installation remains predominantly manual due to the intricate and fine-grained nature of connector mating.

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.

Video about Experiments

AI-based Framework for Visuotactile Connector Mating

An initial robot program is created by the programmer using industry-standard tools ARTM (left). During the ramp-up phase, the robot executes the program repeatedly with varying search parameterizations. The resulting dataset is used to train MuTT, a predictive visuotactile model of the robot and environment dynamics (center). This model then serves as a predictor for the first-order optimizer SPI, which generates parameters that optimize the program for robust and fast connector mating given the observed process variance and the current environment (right).

BibTeX

@misc{kienle2025aibasedframeworkrobustmodelbased,
      title={AI-based Framework for Robust Model-Based Connector Mating in Robotic Wire Harness Installation}, 
      author={Claudius Kienle and Benjamin Alt and Finn Schneider and Tobias Pertlwieser and Rainer Jäkel and Rania Rayyes},
      year={2025},
      eprint={2503.09409},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2503.09409}, 
}

Acknowledgement

This work was supported by the German Ministry for Economic Affairs and Climate Action under grant 13IK026A.