Papers
35
Total Citations
1,189
H-Index
16
About
Michael Neunert is a prominent robotics and control researcher whose work bridges optimal control, trajectory optimization, and reinforcement learning. He has made substantial contributions to real-time nonlinear model predictive control, most notably through his unified framework combining trajectory optimization and tracking control, which has garnered over 200 citations and demonstrated that sophisticated control can be achieved at practical computational speeds. His influential work on trajectory optimization for legged robots — including automatic gait discovery for quadrupeds and foothold optimization — has helped redefine how robots plan and execute complex dynamic movements through contact-rich environments, collectively accumulating hundreds of citations across multiple papers. Beyond classical control, Neunert has made notable strides in reinforcement learning, particularly in sparse-reward settings. His Scheduled Auxiliary Control (SAC-X) framework, cited over 150 times, enables agents to learn complex behaviors from scratch using auxiliary tasks — a meaningful advance for real-world robot learning. His later contributions to offline reinforcement learning further address the practical challenges of deploying RL in robotics without live data collection. Adding to his community impact, Neunert developed the open-source Control Toolbox, providing the robotics community with efficient tools for modeling, optimization, and control. His body of work reflects a researcher deeply committed to making advanced control and learning methods practically deployable on real robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2Learning by Playing - Solving Sparse Reward Tasks from Scratch155 citations · 2018
- 3
- 4
- 5
- 6
- 7
- 8
- 9Online walking motion and foothold optimization for quadruped locomotion46 citations · 2017
- 10