The Effect of Leg Design on Robotic Running Gaits

Advances in legged robotics have led to the development of robots capable of running over rough terrain. Edubot, one such hexapedal robot, is a biologically inspired runner that mimics the cockroach in its alternating tripod gait and compliant legs. Most animals’ running gaits can be approximated by the Spring-Loaded Inverted Pendulum (SLIP) model, which treats an animal as a point mass on a linear spring. The SLIP model has been demonstrated to accurately model the center of mass motion of running animals, as well as the ground reaction forces associated with their gaits. While Edubot’s running behavior can also be viewed in terms of the SLIP model, the linear spring in the model fails to capture the complexity of the compliant C-shaped legs of the robot. These legs have shown superior performance over all previous designs. In this paper we attempt to isolate the characteristics of the C-legs and study their individual and combined effect on performance in hopes of understanding how to design better legs. We do this by decoupling and mathematically modeling the spring rest length and the effective stiffness of the leg as it rolls through stance. Using a SLIP model modified with our equations, we show evidence that both a decreasing stiffness and an increasing spring rest length during stance are important contributors to performance. We then describe legs designed to test our model’s predictions.