Controlling a system with redundant degrees of freedom: transition from standing to walking.

In earlier investigations Lévy and Cruse (J Comp Physiol 194:719-733, 2008; J Comp Physiol 194:735-750, 2008) asked the question that how the nervous system of a stick insect, Carausius morosus, chooses a solution from an abundance of possibilities to solve the task of distributing its body weight onto its six legs, i.e., the torques that are produced by the 18 joints (three per leg). This paper concentrates on the transition from standing to walking using semi-parametrical regression models. Results presented are basically descriptive and do not deal yet with underlying mechanisms. Based on torque changes, the first swing movement is initiated at about 0.3 s before the first leg takes off. The way torques change depends on the joint type, on whether the leg is going to swing and on the number of legs swinging, but not on the leg type, not on the torques size and not on the stepping configuration. Furthermore, the more legs are lifted off to begin a swing movement the higher are the torques produced. The results might be interpreted such that legs starting a swing movement develop, before lift off, levating torques to overcome the adhesive forces produced by the tarsi, whereas the other legs develop depressing torques to maintain the posture.