In this project, the trajectory control of a muscle-actuated robotic manipulator is designed using NeuraBASE. This approach does not require a mathematical model for the control system. Its online training feature allows it to continuously adapt to a changing environment. The successive interval halving method is used to define the trajectory of the muscle arm and the NeuraBASE controller learns to follow the desired trajectory using reinforcement learning.

https://neuramatix.com/projects/videos/ArtificialMuscle.mp4