On extracting design principles from biology: II. Case study-the effect of knee direction on bipedal robot running efficiency.

Comparing the leg of an ostrich to that of a human suggests an important question to legged robot designers: should a robot's leg joint bend in the direction of running ('forwards') or opposite ('backwards')? Biological studies cannot answer this question for engineers due to significant differences between the biological and engineering domains. Instead, we investigated the inherent effect of joint bending direction on bipedal robot running efficiency by comparing energetically optimal gaits of a wide variety of robot designs sampled at random from a design space. We found that the great majority of robot designs have several locally optimal gaits with the knee bending backwards that are more efficient than the most efficient gait with the knee bending forwards. The most efficient backwards gaits do not exhibit lower touchdown losses than the most efficient forward gaits; rather, the improved efficiency of backwards gaits stems from lower torque and reduced motion at the hip. The reduced hip use of backwards gaits is enabled by the ability of the backwards knee, acting alone, to (1) propel the robot upwards and forwards simultaneously and (2) lift and protract the foot simultaneously. In the absence of other information, designers interested in building efficient bipedal robots with two-segment legs driven by electric motors should design the knee to bend backwards rather than forwards. Compared to common practices for choosing robot knee direction, application of this principle would have a strong tendency to improve robot efficiency and save design resources.