The Use of Reinforcement Learning in the Task of Moving Objects with the Robotic Arm

  • Ermek E. AitygulovEmail author
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11866)


The article describes the task of controlling a robotic arm to transfer objects in front of it. To select actions, the reinforcement learning algorithm is used. In conclusion, there are presented the results of experiments in the Gazebo simulation environment with two different inputs: either with information about the position of the hand and the object, or with information about the position of the hand and the image with the camera.


Robotic arm Reinforcement learning Object manipulation 


  1. 1.
    Albers, A., Yan, W., Frietsch, M.: Application of Reinforcement Learning for a 2-DOF Robot Arm Control, November 2009Google Scholar
  2. 2.
    James, S., Johns, E.: 3D Simulation for Robot Arm Control with Deep Q-Learning. 2016Google Scholar
  3. 3.
    Kakade, S.: A natural policy gradient (2002)Google Scholar
  4. 4.
    Schulman, J., Levine, S., Moritz, P., Jordan, M., Abbeel, P.: Trust region policy optimization (2015)Google Scholar
  5. 5.
    Gu, S., Holly, E., Lillicrap, T., Levine, S.: Deep reinforcement learning for robotic manipulation with asynchronous off-policy update (2016)Google Scholar
  6. 6.
    Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction (1998)CrossRefGoogle Scholar
  7. 7.
    Watkins, C.J.C.H.: Learning from delayed rewards (1989)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Moscow Institute of Physics and TechnologyMoscowRussia

Personalised recommendations