Towards Rich Motion Skills with the Lightweight Quadruped Robot Serval - a Design, Control and Experimental Study Peter Eckert 1 , Anja E. M. Schmerbauch 1,2 , Tomislav Horvat 1 , Katja S¨ ohnel 3 , Martin S. Fischer 3 , Hartmut Witte 2 , and Auke J. Ijspeert 1 1 Biorobotics Laboratory, ´ Ecole Polytechnique F´ ed´ erale de Lausanne, Switzerland 2 Fachgebiet Biomechatronik, Technische Universit¨ at Ilmenau, Germany 3 Institut f¨ ur Zoologie und Evolutionsforschung, Friedrich-Schiller-Universit¨ at Jena, Germany Abstract. Bio-inspired robotic designs introducing and benefiting from morphological aspects present in animals allowed the generation of fast, robust and energy efficient locomotion. We used engineering tools and interdisciplinary knowledge transferred from biology to build low-cost robots able to achieve a certain level of versatility. Serval, a compliant quadruped robot with actuated spine and high range of motion in all joints was developed to address the question of what mechatronic complexity is needed to achieve rich motion skills. In our experiments, the robot presented a high level of versatility (number of skills) at medium speed, with a minimal control effort and, in this article, no usage of its spine. Implementing a basic kinematics-duplication from dogs, we found strengths to emphasize, weaknesses to correct and made Serval ready for future attempts to achieve more agile locomotion. In particular, we investigated the following skills: trot, bound (crouched), sidestep, turn with a radius, ascend slopes including flat ground transition, perform single and double step-downs, fall, trot over bumpy terrain, lie/sit down, and stand up. 1 Introduction When we think about animals, we see many species moving dynamically in their natural habitats. While one also observes static behaviors, being in motion is characteristic for most of animal life. The variety of motion skills present in a single species is vast and more refined than what robotics researchers were able to achieve so far with technology. Two aspects that are especially striking in this context are versatility and agility. Presented originally in [1], agility is defined as follows: ”Agility is representing a previously acquired and size dependent set of locomotion skills, executed in a precise, fast and ideally reflexive manner to an outside stimulus.”. Comparing this definition to the one of versatility from the Oxford Dictionary (en.oxforddictionaries.com): ”Ability to adapt or be adapted to many functions or activities” a new definition of agility as fast versatility or fast execution of rich motion skills may be called for. In consequence, one way to reach agility, could be to first achieve versatile behavior followed by an increase in its execution speed. Machines with relatively simple underlying principles (e.g., cars or bikes) can move very well in our environment and navigate even through difficult terrains. In legged robotics on the other hand, whose motivation is high adaptability to uneven or difficult terrains [2], such fast and reliable locomotion is yet to be achieved. It is still unclear, although explored in many laboratories all over the world, what level of mechatronic complexity is minimally needed and sufficient to realize agile motion. Two approaches, often described and used in control, are templates and anchors [3], both incorporating information gained from observation and analysis of animals. Hereby, a template, following the