The neural control of locomotion in a quadrupedal robot

The authors consider whether there is any essential functional difference between the neural control systems involved in instinctive and learned walking. Is there incomplete overlap between the solutions accessible to evolutionary methods (or their synthetic counterparts, genetic algorithms) and those which can be reached by learning methods (or their synthetic counterparts, such as back propagation of error in artificial neural networks)? They consider how these solutions relate to the range of neurally-implemented control systems which an engineer might design. The paper first describes an experimental environment within which these questions may be examined, then examines some theoretical issues, and finally reports their results. The study is based around a quadrupedal walker