TR2020-144

Deep Reactive Planning in Dynamic Environments


    •  Ota, K., Jha, D.K., Onishi, T., Kanezaki, A., Yoshiyasu, Y., Mariyama, T., Nikovski, D.N., "Deep Reactive Planning in Dynamic Environments", Conference on Robot Learning (CoRL), Kober, Jens and Ramos, Fabio and Tomlin, Calire, Eds., November 2020, pp. 1943-1957.
      BibTeX TR2020-144 PDF Video
      • @inproceedings{Ota2020nov2,
      • author = {Ota, Kei and Jha, Devesh K. and Onishi, Tadashi and Kanezaki, Asako and Yoshiyasu, Yusuke and Mariyama, Toshisada and Nikovski, Daniel N.},
      • title = {Deep Reactive Planning in Dynamic Environments},
      • booktitle = {Conference on Robot Learning (CoRL)},
      • year = 2020,
      • editor = {Kober, Jens and Ramos, Fabio and Tomlin, Calire},
      • pages = {1943--1957},
      • month = nov,
      • publisher = {Proceedings of Machine Learning Research},
      • url = {https://www.merl.com/publications/TR2020-144}
      • }
  • MERL Contacts:
  • Research Area:

    Robotics

Abstract:

The main novelty of the proposed approach is that it allows a robot to learn an end-to-end policy which can adapt to changes in the environment during execution. While goal conditioning of policies has been studied in the RL literature, such approaches are not easily extended to cases where the robot’s goal can change during execution. This is something that humans are naturally able to do. However, it is difficult for robots to learn such reflexes (i.e., to naturally respond to dynamic environments), especially when the goal location is not explicitly provided to the robot, and instead needs to be perceived through a vision sensor. In the current work, we present a method that can achieve such behavior by combining traditional kinematic planning, deep learning, and deep reinforcement learning in a synergistic fashion to generalize to arbitrary environments. We demonstrate the proposed approach for several reaching and pick-and-place tasks in simulation, as well as on a real system of a 6-DoF industrial manipulator.

 

  • Related Video

  • Related Publication

  •  Ota, K., Jha, D.K., Onishi, T., Kanezaki, A., Yoshiyasu, Y., Mariyama, T., Nikovski, D.N., "Deep Reactive Planning in Dynamic Environments", arXiv, October 2020.
    BibTeX arXiv
    • @article{Ota2020oct,
    • author = {Ota, Kei and Jha, Devesh K. and Onishi, Tadashi and Kanezaki, Asako and Yoshiyasu, Yusuke and Mariyama, Toshisada and Nikovski, Daniel N.},
    • title = {Deep Reactive Planning in Dynamic Environments},
    • journal = {arXiv},
    • year = 2020,
    • month = oct,
    • url = {https://arxiv.org/pdf/2011.00155.pdf}
    • }