Michael Neunert neunertm (at) ethz (dot) ch

Michael Neunert is a PhD student at the Agile & Dexterous Robotics Lab at ETH Zurich. His research is focused on control, planning, state estimation and machine learning for robots. Michael is eager to find general methods that apply to a broad range of highly dynamic, underactuated robots such as legged robots, multicopters, robotic arms and balancing robots.

Michael received both a Bachelor and a Master degree in Mechanical Engineering from ETH Zurich. During this period, Michael was a founding member of project Rezero http://www.rezero.ethz.ch, developing the world's most agile ball balancing robot. For his master thesis, he joined the EU FP7 project V-Charge http://www.v-charge.eu/ which successfully demonstrated autonomous driving with low-cost sensors. His thesis “Online Localization for an Autonomous Car Using Multiple Cameras” has been awarded with the “Johann Puch Innovation Award” by Magna Steyr.

In his ongoing research, Michael is developing high performance optimal control algorithms for trajectory optimization and control. These algorithms enable long horizon, model predictive control using complete, non-simplified models. The capabilities of these algorithms have been successfully demonstrated on multicopters, mobile manipulators and other underactuated, agile robots. Currently, Michael is working on bringing these algorithms to legged systems with many degrees of freedom.

[1] Michael Neunert, Michael Blösch, Jonas Buchli (2015). An Open Source, Fiducial Based, Visual-Inertial State Estimation System. arXiv, 1507.02081 link

[2] Diego Pardo, Lukas Möller, Michael Neunert, Alexander W. Winkler, Jonas Buchli (2015). Evaluating direct transcription and nonlinear optimization methods for robot motion planning. arXiv, 1504.05803. http://arxiv.org/abs/1504.05803

[3] Cedric de Crousaz, Farbod Farshidian, Michael Neunert, Jonas Buchli (2015). Unified Motion Control for Dynamic Quadrotor Maneuvers Demonstrated on Slung Load and Rotor Failure Tasks. In IEEE International Conference on Robotics and Automation, pp. 2223-2229. pdf

[4] Farbod Farshidian, Michael Neunert, Jonas Buchli (2014). Learning of Closed-Loop Motion Control. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) pdf

[5] Michael Neunert, Farbod Farshidian, Jonas Buchli (2014). Adaptive Real-time Nonlinear Model Predictive Motion Control. In IROS 2014 Workshop on Machine Learning in Planning and Control of Robot Motion http://www.adrl.ethz.ch/archive/p_14_mlpc_mpc_rezero.pdf

RCARS - An open source, fiducial based, visual-inertial state estimation system RCARS, publication see [1]