Legged robots offer mobility and agility in uncharted terrains. Tracking is central to legged operations and has traditionally been performed using inertial measurement techniques. The discontinuous foot fall patterns and flight phases that yield this unrivaled mobility serve to limit the motion measurement. In particular, the severe impact from repeated leg landings results in an excessive accumulation of drift. Ground range measurements, amongst several others, are robust to this drift yet

more »

... limited in application due to their low-bandwidth and variability to ground conditions. This paper outlines the attitude estimation problem for legged locomotion, extends the ground range measurement method as an update to an inertial sensing approach, introduces the use of a hybrid estimator based on the flightphases of the legged motion, and shows preliminary results this combined approach. Based on an Extended Kalman Filter, the method takes advantage of the mostly ballistic nature of the flight phases in dynamic locomotion. Initial results indicate that the method provides fast update rates yet controls drift. In single leg experiments, which were conducted using low-cost sensing hardware, this method had an RMS error of <0.7 • , which was a third that of the next comparable approach. Together this provides rapid, robust estimates of flight phases and attitude necessary for extended dynamic legged operations.