Abstract
Robotic peg-in-hole task has always attracted researchers’ attention. With the development of real-time sensors and machine learning algorithms, collaborative robots are now having potential to insert tiny and delicate components of digital products. Due to grasping error, the absolute position of the peg would not be calculated directly by forward kinematics, but through high-resolution sensors. However, for each single modality, such as RGB-D image and proprioception, has its own limitation during the insertion process. Camera cannot provide accurate information when the peg is closed to the target, while force/torque sensor is entirely blind before contact status begin. This paper used multimodal fusion method to utilize all the valuable information from multiple sensors. Representation cores from multimodal data were trained to forecast relative position between the peg and hole. Reinforcement learning network was then able to use the relative position to generate appropriate action of the robot. This paper verified the above algorithms through USB-C insertion experiments in ROS-Gazebo simulation.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Sahin, C., Kim, T.-K.: Recovering 6D object pose: a review and multi-modal analysis. In: Leal-Taixé, L., Roth, S. (eds.) Computer Vision – ECCV 2018 Workshops. LNCS, vol. 11134, pp. 15–31. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-11024-6_2
Kapusta, A., Park, D., Kemp, C.C.: Task-centric selection of robot and environment initial configurations for assistive tasks. In: IEEE International Conference on Intelligent Robots and Systems (IROS), pp. 1480–1487 (2015)
Rozo, L., Bruno, D., Calinon, S., Caldwell, D.G.: Learning optimal controllers in human-robot cooperative transportation tasks with position and force constraints. In: IEEE International Conference on Intelligent Robots and Systems (IROS), pp. 1024–1030 (2015)
Bartolozzi, C., Natale, L., Nori, F., Metta, G.: Robots with a sense of touch. Nat. Mater. 15(9), 921–925 (2016)
Liang, P., Ge, L., Liu, Y., Zhao, L., Li, R., Ke, W.: An augmented discrete-time approach for human-robot collaboration. Discret. Dyn. Nat. Soc. (2016)
Xue, T., et al.: bayesian grasp: robotic visual stable grasp based on prior tactile knowledge (2019). http://arxiv.org/abs/1905.12920
Song, H.C., Kim, Y.L., Song, J.B.: Automated guidance of peg-in-hole assembly tasks for complex-shaped parts. In: IEEE International Conference on Intelligent Robots and Systems (IROS), pp. 4517–4522 (2014)
Lee, M.A., et al.: Making sense of vision and touch: learning multimodal representations for contact-rich tasks. IEEE Trans. Robot. 36(3), 582–596 (2019)
Wang, Z., Yang, X., Hu, H., Lou, Y.: Actor-critic method-based search strategy for high precision peg-in-hole tasks. In: 2019 IEEE International Conference on Real-Time Computing and Robotics (RCAR), pp. 458–463 (2019)
Scherzinger, S., Roennau, A., Dillmann, R.: Contact skill imitation learning for robot-independent assembly programming. In: IEEE International Conference on Intelligent Robots and Systems (IROS), pp. 4309–4316 (2019)
Sidhik, S.: panda_simulator: Gazebo simulator for Franka Emika Panda robot supporting sim-to-real code transfer. Zenodo (2020)
Gu, S., Lillicrap, T., Sutskever, I., Levine, S.: Continuous deep q-learning with model-based acceleration. In: 33rd International Conference on Machine Learning (ICML), vol. 6, pp. 4135–4148 (2016). http://arxiv.org/abs/1603.00748
Acknowledgment
This work was supported partially by the NSFC-Shenzhen Robotics Basic Research Center Program (No. U1713202) and partially by the Shenzhen Science and Technology Program (No. JSGG20191129114035610).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Li, A., Liu, R., Yang, X., Lou, Y. (2021). Reinforcement Learning Strategy Based on Multimodal Representations for High-Precision Assembly Tasks. In: Liu, XJ., Nie, Z., Yu, J., Xie, F., Song, R. (eds) Intelligent Robotics and Applications. ICIRA 2021. Lecture Notes in Computer Science(), vol 13013. Springer, Cham. https://doi.org/10.1007/978-3-030-89095-7_6
Download citation
DOI: https://doi.org/10.1007/978-3-030-89095-7_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-89094-0
Online ISBN: 978-3-030-89095-7
eBook Packages: Computer ScienceComputer Science (R0)