Reward-modulated Hebbian plasticity as leverage for partially embodied control in compliant robotics
- Author
- Jeroen Burms (UGent) , Ken Caluwaerts and Joni Dambre (UGent)
- Organization
- Abstract
- In embodied computation (or morphological computation), part of the complexity of motor control is offloaded to the body dynamics. We demonstrate that a simple Hebbian-like learning rule can be used to train systems with (partial) embodiment, and can be extended outside of the scope of traditional neural networks. To this end, we apply the learning rule to optimize the connection weights of recurrent neural networks with different topologies and for various tasks. We then apply this learning rule to a simulated compliant tensegrity robot by optimizing static feedback controllers that directly exploit the dynamics of the robot body. This leads to partially embodied controllers, i.e., hybrid controllers that naturally integrate the computations that are performed by the robot body into a neural network architecture. Our results demonstrate the universal applicability of reward-modulated Hebbian learning. Furthermore, they demonstrate the robustness of systems trained with the learning rule. This study strengthens our belief that compliant robots should or can be seen as computational units, instead of dumb hardware that needs a complex controller. This link between compliant robotics and neural networks is also the main reason for our search for simple universal learning rules for both neural networks and robotics.
- Keywords
- compliant robotics, Hebbian plasticity, morphological computation, tensegrity, recurrent neural networks, LEARNING RULE, TENSEGRITY, NETWORKS, NOISE, MODEL
Downloads
-
fnbot-09-00009.pdf
- full text
- |
- open access
- |
- |
- 7.51 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-6960324
- MLA
- Burms, Jeroen, et al. “Reward-Modulated Hebbian Plasticity as Leverage for Partially Embodied Control in Compliant Robotics.” FRONTIERS IN NEUROBOTICS, edited by Andrea Soltoggio, vol. 9, no. 9, Frontiers Media SA, 2015, doi:10.3389/fnbot.2015.00009.
- APA
- Burms, J., Caluwaerts, K., & Dambre, J. (2015). Reward-modulated Hebbian plasticity as leverage for partially embodied control in compliant robotics. FRONTIERS IN NEUROBOTICS, 9(9). https://doi.org/10.3389/fnbot.2015.00009
- Chicago author-date
- Burms, Jeroen, Ken Caluwaerts, and Joni Dambre. 2015. “Reward-Modulated Hebbian Plasticity as Leverage for Partially Embodied Control in Compliant Robotics.” Edited by Andrea Soltoggio. FRONTIERS IN NEUROBOTICS 9 (9). https://doi.org/10.3389/fnbot.2015.00009.
- Chicago author-date (all authors)
- Burms, Jeroen, Ken Caluwaerts, and Joni Dambre. 2015. “Reward-Modulated Hebbian Plasticity as Leverage for Partially Embodied Control in Compliant Robotics.” Ed by. Andrea Soltoggio. FRONTIERS IN NEUROBOTICS 9 (9). doi:10.3389/fnbot.2015.00009.
- Vancouver
- 1.Burms J, Caluwaerts K, Dambre J. Reward-modulated Hebbian plasticity as leverage for partially embodied control in compliant robotics. Soltoggio A, editor. FRONTIERS IN NEUROBOTICS. 2015;9(9).
- IEEE
- [1]J. Burms, K. Caluwaerts, and J. Dambre, “Reward-modulated Hebbian plasticity as leverage for partially embodied control in compliant robotics,” FRONTIERS IN NEUROBOTICS, vol. 9, no. 9, 2015.
@article{6960324,
abstract = {{In embodied computation (or morphological computation), part of the complexity of motor control is offloaded to the body dynamics. We demonstrate that a simple Hebbian-like learning rule can be used to train systems with (partial) embodiment, and can be extended outside of the scope of traditional neural networks. To this end, we apply the learning rule to optimize the connection weights of recurrent neural networks with different topologies and for various tasks. We then apply this learning rule to a simulated compliant tensegrity robot by optimizing static feedback controllers that directly exploit the dynamics of the robot body. This leads to partially embodied controllers, i.e., hybrid controllers that naturally integrate the computations that are performed by the robot body into a neural network architecture. Our results demonstrate the universal applicability of reward-modulated Hebbian learning. Furthermore, they demonstrate the robustness of systems trained with the learning rule. This study strengthens our belief that compliant robots should or can be seen as computational units, instead of dumb hardware that needs a complex controller. This link between compliant robotics and neural networks is also the main reason for our search for simple universal learning rules for both neural networks and robotics.}},
author = {{Burms, Jeroen and Caluwaerts, Ken and Dambre, Joni}},
editor = {{Soltoggio, Andrea}},
issn = {{1662-5218}},
journal = {{FRONTIERS IN NEUROBOTICS}},
keywords = {{compliant robotics,Hebbian plasticity,morphological computation,tensegrity,recurrent neural networks,LEARNING RULE,TENSEGRITY,NETWORKS,NOISE,MODEL}},
language = {{eng}},
number = {{9}},
pages = {{14}},
publisher = {{Frontiers Media SA}},
title = {{Reward-modulated Hebbian plasticity as leverage for partially embodied control in compliant robotics}},
url = {{http://dx.doi.org/10.3389/fnbot.2015.00009}},
volume = {{9}},
year = {{2015}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: