Dr. Gavin Miller has been investigating snake robots for over ten years. His first models, S1 and S2, use one servo at each joint. S3 used a pair of servos to create a sidewinding motion along with the more common horizontal undulations. S4 was an attempt to power the snake with just one motor, but ran into problems because of high torque in the u-joints. S5 was the first to have all CNC parts, which made for a more compact package. The larger snake did require the use of a second microprocessor. S6 was another attempt at a single motor snake, this time using a complicated gear train. It still had problems. S7 is the most recent prototype and is an attempt at rectilinear motion like that of a python (undulations in the direction of movement).
[thanks joeboy]
