(logo by Janette Forte)
This is the official site of the UT Austin Villa 3D Simulation team from the Department of Computer Science at the University of Texas at Austin.

This page provides information on evolutionary optimization of walk engine controllers and robot morphology for bipedal running.


Evolutionary Optimization of Walk Engine Controllers and Robot Morphology for Bipedal Running
Download video: mp4

We perform optimization in two different simulators: SimSpark and Gazebo. Both simulators use the Open Dynamics Engine (ODE) library for their realistic simulation of rigid body dynamics with collision detection and friction. The robots in the simulators are modeled after the humanoid Nao robot, which has a height of about 57 cm and a mass of 4.5 kg. Each robot has 22 degrees of freedom: six in each leg, four in each arm, and two in the neck. In total 27 parameters for an omnidirectional walk engine controller were optimized with the CMA-ES algorithm to produce fast and stable running behavior. Additionally, in the SimSpark simulator, we optimized six anchor joint positions of the robot's body: the X,Y,Z offset values for the robot's hinge joints in its ankles, knees, and hips for a total of 18 additional parameter values.

For any questions, please contact Patrick MacAlpine.