Towards a Comparative Measure of Legged Agility

Abstract

We introduce an agility measure enabling the comparison of two very different leaping-from-rest transitions by two comparably powered but morphologically different legged robots. We use the measure to show that a flexible spine outperforms a rigid back in the leaping-from-rest task. The agility measure also sheds light on the source of this benefit: core actuation through a sufficiently powerful parallel elastic actuated spine outperforms a similar power budget applied either only to preload the spine or only to actuate the spine during the leap, as well as a rigid backed configuration of the identical machine. Abstract. We introduce an agility measure enabling the comparison of two very di↵erent leaping-from-rest transitions by two comparably powered but morphologically di↵erent legged robots. We use the measure to show that a flexible spine outperforms a rigid back in the leapingfrom-rest task. The agility measure also sheds light on the source of this benefit: core actuation through a su ciently powerful parallel elastic actuated spine outperforms a similar power budget applied either only to preload the spine or only to actuate the spine during the leap, as well as a rigid backed configuration of the identical machine. Disciplines Electrical and Computer Engineering | Engineering | Systems Engineering

Core Actuation Promotes Self-manipulability on a Direct-Drive Quadrupedal Robot

“…5. Leaping from the ground with and without spine bending using an otherwise identical feed-forward control scheme shows that the spine motors add on average 5.7 J to the body energy [13] (discounting the 0.5 J stored in initial spine elastic potential energy). The body energy added is calculated by subtracting the energy at the leap height apex-indicated by a vertical tick in the sample energetic traces shown in the right figure-from the starting energy.…”

“…These energetic results were calculated from the extrinsic body energy of the robot, the sum of the mass center's kinetic and gravitational potential energy relative to a simplified Lagrangian model, as documented in [13]. Leaping aided by spine bending added an average of 22.8 ± 0.5 J to the body (an average of 22.3 J when discounting the elastic potential energy separately measured to be stored in the spine's fiberglass plate bending) and leaping with an identical strategy but without bending the spine added an average of 16.6 ± 0.7 J to the body, 6.2 J less than with spine bending.…”

“…Other experiments have suggested increased running speed [11] and gait transition stability [12]. Leaping from a crouched position with a parallel elastic-actuated spine was shown to increase leap energy in [13]. However, much more work is needed to provide designers with the quantifiable utility of core actuation or lack thereof, especially for the less-studied application of traversing irregular terrain.…”

Empirical validation of a spined sagittal-plane quadrupedal model

“…The design of the power-autonomous MIT Cheetah utilizing core actuation is presented in [15] but only simulation data Electrical and Systems Engineering Department, University of Pennsylvania, 200 South 33rd Street, Philadelphia, PA 19104 {jdup,kod}@seas.upenn.edu appears to be given. Other spined platforms exist such as the Canid robot but have only been documented executing non-steady-state tasks [16], [17], [18].…”

Leg design for energy management in an electromechanical robot

“…Specifically, the stance event of this dynamical system can be used to pull both efficiency [5] and agility [6] criteria back into the design space of the mechanism.…”