Resumen
This article has surveyed the efforts of the authors over the past nine years to incorporate some of the principles of insect walking into the design and control of legged robots. This work has highlighted the robotic applications of several general principles of insect walking, including highly distributed control; synergistic combination of local reflexes with distributed gait control; matching the mechanical properties of the body to the demands of the terrain and the design of the controller. The application of these principles to the design and control of legged robots has resulted in devices that can effectively move over complex terrain with only very modest computational resources using highly distributed control systems that are very robust. Insect legs have many more degrees of freedom, including both flexible segments and claws that allow them to grip unstable surfaces and to walk along steep inclines or even on inverted surfaces. In addition, muscle is a far more powerful and efficient actuator than the best electric motors currently available. Insect legs are populated with hundreds of sensors that provide far more information about the positions, velocities, and accelerations of each joint and the stresses experienced by each segment than current legged robots have access to. Antennae, which are also densely populated with a variety of sensory structures, are also used by insects to adjust their leg movements to the terrain ahead. Most importantly of all, insect nervous systems integrate this vast collection of sensory feedback to flexibly reconfigure the many available degrees of freedom on a step-by-step basis, producing steady locomotion under an amazingly diverse set of conditions.