Additional DOFs and Sensors for Bio-inspired Locomotion: Towards Active Spine, Ankle Joints, and Feet for a Quadruped Robot
In Proceedings of the International Symposium on Artificial Intelligence, Robotics and Automation in Space (i-SAIRAS 2012), (iSAIRAS-12), 04.9.-06.9.2012, Turin, o.A., Sep/2012.
This paper describes the development of a mobile walking robot for rough terrain e.g. crater environments. The focus lies on the development of a robotic research platform incorporating biologically inspired structural components which, if employed on the robotic system, effectively improve the locomotion and mobility characteristics. In order to achieve the above mentioned goal, an improved perception of the system state and its environment is needed. The intelligent structures (iStruct) contain a variety of functions which cannot only extend the already existing locomotion behaviors of robots, but also permit further relevant applications like the contemporaneous use of the structure as carrier and sensor system. The underlying ideas how to increase the mobility of robotic systems by using biological concepts regarding control aspects and mechanical design are described in this paper. In robotics, a rigid torso is often employed. Within the proposed system, the connection between the shoulder and hip is an actuated, artificial spinal structure with six degrees of freedom (DoFs). In
order to increase the robustness of the systems locomotion in terms of traction and stability, a foot-like structure equipped with several sensors (49 x pressure sensors, a 6-DoF force/torque sensor, an acceleration sensor, a distance sensor, and temperature sensors) was developed. Using this structure, the robotic system obtains a better perception of its environment and has therefore e.g. the possibility to control the interaction between its planar sole and the ground. It is pointed out, how the general idea behind predictive control or the efference copy principle can be applied to a robotic system. The usage of printed circuit boards within each of the structures allows a distributed control; the structures are as self-contained as possible in terms of sensing, sensor preprocessing, motor control, and communication. In addition to this, results with the assembled mechanical parts regarding sensor fusion, local control loops, and maneuverability for the lower leg, the foot, and the spine are shown.