Abstract
This paper proposes a footstep planning method for centipede-like multi-legged robots from tactile information of antennal sensor. Centipede-like multi-legged walking robots are controlled based on Follow-the-Contact-Point (FCP) gait control. Using FCP gait control, merely planning and allocating the contact points of legs of a head segment adequately allows the robots to change the moving direction and also step on rough terrain. To plan the contact points, two antennal sensors are added to the head of the centipede-like multi-legged walking robot. When the antenna makes contact with an obstacle, the collision position of the antennal sensor is recorded to make a temporal contact map. An optimal target contact point of head legs is selected from this contact map. Finally, in order to show the availability of the proposed method, the results of physical simulation are shown.
