breveWalker is an experiment in artificial intelligence and robotics which uses a genetic algorithm to teach a physically simulated legged robot how to walk.  Starting with completely random behaviors, the system uses evolution to discover walking strategies which carry the robot the farthest.


Starting with completely randomized data, a genetic algorithm tests potential solutions one by one, breeding and mutating those that preform well and discarding those that do not.  Over time, the randomized data evolves into robust solutions to problem that the genetic algorithm is trying to solve.


breveWalker is based on the breve simulation environment.


Contact the author of breveWalker for support.

Get it for your iPhone in the iTunes App Store – $2.99http://artificial.com/a/bwhttp://artificial.com/a/bwshapeimage_4_link_0shapeimage_4_link_1