Advanced search
1 file | 460.07 KB Add to list

Towards accelerated robotic deployment by supervised learning of latent space observer and policy from simulated experiments with expert policies

Olivier Algoet (UGent) , Tom Lefebvre (UGent) and Guillaume Crevecoeur (UGent)
Author
Organization
Abstract
Up until today robotic tasks in highly variable environments remain very difficult to solve. We propose accelerated robotic deployment through task solving on low-level sensor data in simulation. A simulation allows for a lot of data, which is usually not available in a real world robotic setup due to cost and feasibility. Solving tasks in simulation is safe and a lot easier due to the huge amount of feedback from virtual sensory data. We present a novel sim2real architecture for converting simulated low level sensor data policies to high level real world policies. After solving a task we let the robot complete it a number of times in simulation using domain randomization, while doing so we save the simulated sensor data corresponding to the real robotic setup and actions taken. Given these sensor data and actions a task specific policy can be trained using our architecture. In this paper we work towards a proof of concept by simulating a simple low cost manipulator in pybullet to pick and place an object based on image observations.
Keywords
Domain Randomization, Sim2real, behavioral cloning, Deep learning, Robotics

Downloads

  • (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 460.07 KB

Citation

Please use this url to cite or link to this publication:

MLA
Algoet, Olivier, et al. “Towards Accelerated Robotic Deployment by Supervised Learning of Latent Space Observer and Policy from Simulated Experiments with Expert Policies.” 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), IEEE, 2020, pp. 1329–34, doi:10.1109/AIM43001.2020.9158908.
APA
Algoet, O., Lefebvre, T., & Crevecoeur, G. (2020). Towards accelerated robotic deployment by supervised learning of latent space observer and policy from simulated experiments with expert policies. 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), 1329–1334. https://doi.org/10.1109/AIM43001.2020.9158908
Chicago author-date
Algoet, Olivier, Tom Lefebvre, and Guillaume Crevecoeur. 2020. “Towards Accelerated Robotic Deployment by Supervised Learning of Latent Space Observer and Policy from Simulated Experiments with Expert Policies.” In 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), 1329–34. New York: IEEE. https://doi.org/10.1109/AIM43001.2020.9158908.
Chicago author-date (all authors)
Algoet, Olivier, Tom Lefebvre, and Guillaume Crevecoeur. 2020. “Towards Accelerated Robotic Deployment by Supervised Learning of Latent Space Observer and Policy from Simulated Experiments with Expert Policies.” In 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), 1329–1334. New York: IEEE. doi:10.1109/AIM43001.2020.9158908.
Vancouver
1.
Algoet O, Lefebvre T, Crevecoeur G. Towards accelerated robotic deployment by supervised learning of latent space observer and policy from simulated experiments with expert policies. In: 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM). New York: IEEE; 2020. p. 1329–34.
IEEE
[1]
O. Algoet, T. Lefebvre, and G. Crevecoeur, “Towards accelerated robotic deployment by supervised learning of latent space observer and policy from simulated experiments with expert policies,” in 2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), Boston, MA, USA, 2020, pp. 1329–1334.
@inproceedings{8681368,
  abstract     = {{Up until today robotic tasks in highly variable environments remain very difficult to solve. We propose accelerated
robotic deployment through task solving on low-level sensor data in simulation. A simulation allows for a lot of data, which is usually not available in a real world robotic setup due to cost and feasibility. Solving tasks in simulation is safe and a lot easier due to the huge amount of feedback from virtual sensory data. We present a novel sim2real architecture for converting simulated low level sensor data policies to high level real world policies. After solving a task we let the robot complete it a number of times in simulation using domain randomization, while doing so we save the simulated sensor data corresponding to the real robotic setup and actions taken. Given these sensor data and actions a task specific policy can be trained using our architecture. In this paper we work towards a proof of concept by simulating a simple low cost manipulator in pybullet to pick and place an object based on image observations.}},
  author       = {{Algoet, Olivier and Lefebvre, Tom and Crevecoeur, Guillaume}},
  booktitle    = {{2020 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM)}},
  isbn         = {{9781728167947}},
  issn         = {{2159-6255}},
  keywords     = {{Domain Randomization,Sim2real,behavioral cloning,Deep learning,Robotics}},
  language     = {{eng}},
  location     = {{Boston, MA, USA}},
  pages        = {{1329--1334}},
  publisher    = {{IEEE}},
  title        = {{Towards accelerated robotic deployment by supervised learning of latent space observer and policy from simulated experiments with expert policies}},
  url          = {{http://doi.org/10.1109/AIM43001.2020.9158908}},
  year         = {{2020}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: