Leveraging reinforcement and curriculum learning for flexible robot-based snap-fit assembly automation

at - Automatisierungstechnik, vol. 73, pp. 331–340

Abstract

Abstract This work proposes a novel approach combining reinforcement learning (RL) and curriculum learning for flexible robot-based snap-fit assembly. The proposed system entails a simulation environment to train RL models for snap-fit connections, adapting part geometries and friction coefficients during training. The results show significant improvements in assembly success rates over the training stages, reaching over 95 % after iterative adjustments. Future work will focus on transferring the trained models to real robots and addressing the Sim2Real-Gap.

Authors 4

1. Josefine Monnet Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University Laboratory for Machine Tools and Production Engineering: Werkzeugmaschinenlabor der RWTH Aachen , Aachen , Germany

2. Emma Heyen Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University Laboratory for Machine Tools and Production Engineering: Werkzeugmaschinenlabor der RWTH Aachen , Aachen , Germany

3. Oliver Petrović Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University Laboratory for Machine Tools and Production Engineering: Werkzeugmaschinenlabor der RWTH Aachen , Aachen , Germany

4. Werner Herfs Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University Laboratory for Machine Tools and Production Engineering: Werkzeugmaschinenlabor der RWTH Aachen , Aachen , Germany