Torque Controlled or Intrinsically Compliant? DLR's Perspective on Robust and Efficient Biped and Quadruped Locomotion

Abstract

Robots are not only machines which are supposed to relieve humans from dangerous or routine work - they are also a scientific endeavour attempting to better understand human and animal motion and intelligence in a synthetizing way, by using the system analytic tools of engineering and computer science. As such, legged robots, in particular humanoids and quadrupeds, attracted a lot of attention and research effort in recent years. The exploding commercial interest in humanoids in the last two years underlines the huge potential of this technology. Walking robots are supposed to closely interact with their human users or to operate in remote, unknown environments - in both cases, robustness is a central issue, as precise mathematical models for the interaction cannot be expected. I will present in this talk two approaches to legged locomotion, which we followed during the last decade at DLR in order to achieve performance and robustness. The first approach, used for the development of the humanoid TORO, leverages the torque-controlled technology initiated at DLR for robot manipulators, which was subsequently commercialized by KUKA, Franka.Emika, Agile Robots and Medtronic. Precise joint torque interfaces allow performant whole-body control and motion planning for bipedal locomotion, also on uneven ground, as well as safe interaction with humans. Controlling motion at low energetic cost, both from mechanical and computational point of view, certainly constitutes one of the major locomotion challenges in biology and robotics. With the design of our experimental elastic quadruped robot Bert we attempt to demonstrate that robots can be designed and controlled to walk highly efficient by exploiting resonance body effects. To do so, however, legged robots need to achieve limit cycle motions of the highly coupled, non-linear body dynamics. This led us to fundamental research on the theory of intrinsic modal oscillations of nonlinear systems as well as on their stabilisation and control. I will present recent results in this direction from my ERC Advanced Grant project M-Runners. Finally, putting the human in the centre of robot development also means going beyond the pure field of engineering and interacting with bio-sciences. I will particularly highlight in this respect the interplay of biomechanics and neuro-control with advanced robotics design and control. Humans can also directly benefit from this research through the development of better human-machine interfaces, robotized medical procedures, and prosthetic and rehabilitation devices which will even more reduce the barrier between humans and robots in the future.