Supervised Representation Learning Towards Generalizable Assembly State Recognition

Tim J. Schoonbeek (Corresponding author), Goutham Balachandran, Hans Onvlee, Tim Houben, Shao-Hsuan Hung, Jacek Kustra, Peter H.N. de With, Fons van der Sommen

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Assembly state recognition facilitates the execution of assembly procedures, offering feedback to enhance efficiency and minimize errors. However, recognizing assembly states poses challenges in scalability, since parts are frequently updated, and the robustness to execution errors remains underexplored. To address these challenges, this letter proposes an approach based on representation learning and the novel intermediate-state informed loss function modification (ISIL). ISIL leverages unlabeled transitions between states and demonstrates significant improvements in clustering and classification performance for all tested architectures and losses. Despite being trained exclusively on images without execution errors, thorough analysis on error states demonstrates that our approach accurately distinguishes between correct states and states with various types of execution errors. The integration of the proposed algorithm can offer meaningful assistance to workers and mitigate unexpected losses due to procedural mishaps in industrial settings.
Original languageEnglish
Article number10694722
Pages (from-to)9915-9922
Number of pages8
JournalIEEE Robotics and Automation Letters
Volume9
Issue number11
Early online date25 Sept 2024
DOIs
Publication statusPublished - Nov 2024

Keywords

  • Representation learning
  • assembly state detection
  • assembly state recognition
  • computer vision for manufacturing
  • deep learning methods

Fingerprint

Dive into the research topics of 'Supervised Representation Learning Towards Generalizable Assembly State Recognition'. Together they form a unique fingerprint.

Cite this