Abstract

Full, R.J. 1993. Integration of individual leg dynamics with whole body movement in arthropod locomotion. In Biological Neural Networks in Invertebrate Neuroethnology and Robots. (eds. R. Beer, R. Ritzmann and T. McKenna). Academic Press. Boston. pp. 3-20.

This publication, which is the lead chapter in the book, serves as the bridge linking neurobiologists to artificial intelligence and robotic engineers. It represents the movement of my work into a new area. In the chapter, I discuss how it is simply inadequate to conclude that one knows how an animal locomotes simply because the pattern of motor neuron output is described. Muscles receiving the same motor pattern may generate energy, transfer energy or absorb energy, all resulting in very different movements. In supporting this argument I discuss several of the design criteria adopted by robotics engineers that now limit speed and maneuverability of robots. These include: 1. motion of the body or center of mass. The center of mass in multi-legged animals does not move smoothly like a wheel, but oscillates. 2. quasi-static versus dynamic stability. Many-legged animals employ dynamic stability which may reduced the degree of control required. 3. variation in individual leg dynamics - similarity in "effective" leg function. Legs of animals function in very similar ways, like pendulums and springs, neither of which is employed by the majority of robots.