Spiking Neural Network Control

Understanding how Spiking Neurons in the Brain Achieve Motor Control

Spiking Neuron Motor Control: Five pools of spiking neurons compute a control transformations that projects specified task forces into the motor torques. This complex and nonlinear transformation---a Jacobian---helped control the robot above in a fluid and perturbation resistant manner.

Our Methods

Understanding how the brain computes complex motor coordination patterns while performing day-to-day motor tasks at a mechanistic level requires being able to build physiologically plausible models that use spiking neurons to achieve motor control. Operational space control theory has successfully generated human-like movements in robots, providing valuable insights into how humans themselves might move.

Our Research Goals

In the near future, we plan to generalize our spiking neural control algorithms to work with human musculoskeletal models. Testing whether our control model can predict neural motor control maps and whole body motions in human subjects promises novel insights for neuroscience.

Our long term goal is to use our control models for improving rehabilitation strategies and neural prosthetic devices.


[2016] Menon. S, Sriram. V, Kumanduri. L, Khatib. O, Boahen. K“ Controlling a Redundant Articulated Robot in Task Space with Spiking Neurons”, The 25th International Conference on Artificial Neural Networks, 2016. Download paper Bibtex

[2014] Menon. S, Fok. S, Neckar. A, Khatib. O, Boahen. K, “Controlling Articulated Robots in Task-Space with Spiking Silicon Neurons”, Proceedings of the IEEE International Conference on Biomedical Robotics and Biomechatronics, 2014 [IEEE] Download paper Bibtex

[2013] Menon. S, Fok. S, Neckar. A, Chavez. K, Aholt. C, Bertics. S, and Boahen. K “Controlling robot dynamics with spiking neurons”, (Demonstration), Neural Information Processing Systems, 2013 [www] Bibtex

    © Samir Menon. CCA 3.0 license.
    Valid HTML and CSS.
    Last updated on May 2nd, 2015