Gorithm Used For

HeadStep here happens to be V isBug21 4], which uses vision within a limited radius (in our case equal to the link length) to produce locally-optimal paths. Note that during the operation the snake does not collect information about the environment and does not remember its previous path. With this choice of a head planning procedure, the whole algorithm can be thought of as an emulation of the behavior of a real snake: the sensing at the head is equivalent to vision, allowing the head to foresee obstacles and locally optimize its motion; the sensing at the rest of the body emulates tactile sensing, allowing the snake to slide along obstacles. (a) (b) (c) (d) (e) (f) Fig. 5. Example of the algorithm's performance for a 20-link manipulator immersed in a complex environment. The head starts at position S and the user selects T as the target. Six snapshots of the motion are shown. Also shown are the trajectoriesof all the joints between the previous and current snapshots.