Michael Neunert

ETH Zurich

Wednesday, July 05, 2017, 14:00

Hörsaal II, Albertstr. 23b, 79104 Freiburg

Robotics holds high promises contributing solutions to our future society’s challenges by taking over dangerous, physically intense work or fill up the lack of skilled labor due to population aging. To fulfill these promises, robots will need to solve increasingly complex tasks requiring more versatile hardware designs with more degrees of freedom. Paired with fast development cycles, this poses a major challenge for robotic motion planning and control.  Ideally, the task or desired motion is specified on a high level and the planners and controllers then reason about the dynamics and kinematics of the robot to find a solution.
We aim at creating a general, transferable motion planning and control framework by using Numerical Optimal Control. We demonstrate the versatility of Numerical Optimal Control on three entirely different classes of robots namely walking, flying and ground robots and underline the potential with hardware experiments on vastly different tasks. We show how Numerical Optimal Control enables legged robots to discover different gait patterns and allows for aggressively flying unmanned aerial vehicles through confined spaces.
Enabled by the efforts of an efficient implementation, we can run Numerical Optimal Control online or even in Model Predictive Control fashion. As such Numerical Optimal Control can exceed the capabilities of classical robot motion planning and control. Furthermore, it provides the user with a high level interface to specify different desired behaviours.