A Distributed Neural Network Architecture for Hexapod Robot Locomotion

This is the most nuts and bolts of randy's paper's that I've read.

Intriguing quote: "In addition, a lesion study of this network demonstrated both its surprising robustness and the subtley of the interaction between its central and peripheral components (Chiel and Beer 1989)."

I want to know about the subtlety of the interaction between its central and peripheral components. This quote betrays a potential weak assumption that both sides of the presumed 'patterns v.s. reflexes' debate are making. Is it safe to associate reflexive with distributed and patterned with centralized?

The image on page 359 (page 4), the model of the network controlling each leg, where does that come from. Is it the distillate of experience getting hexapod robots wrong or is it the distillate of the 'neuroethology'.

This paper is also where Randy describe the emergence of and solution to desynchronization of the opposite sets of legs and where he provides a take on Brook's work with Genghis, the predecessor to Hannibal. Is the desynchronization here the same one that Brook's student Cynthia described encountering in A Comparison of Three Insect Inspired Locomotion Controllers