Real-Time Whole-Body MPC for Legged Robots with MuJoCo

We present a series works showing simple model-predicitve controls algorithms such as MPPI and iLQR can be surprisingly effective at enabling a variety of whole-body control behaviors on high dimensional legged robtots (i.e. quadrupeds and humanoids) in the real world.

Checkout our websites:

mppi
ilqr

Related Papers

2025 March	Whole-Body Model-Predictive Control of Legged Robots with MuJoCo John Zhang, Taylor Howell, Zeji Yi, Chaoyi Pan, Guanya Shi, Quannan Qu, Tom Erez, Yuval Tassa, and Zac Manchester arxiv (In Review)
2025 May	Real-Time Whole-Body Control of Legged Robots with Model-Predictive Path Integral Control Juan Alvarez Padilla, John Zhang, Sofia Kwok, John M. Dolan, and Zac Manchester IEEE International Conference on Robotics and Automation (ICRA). Atlanta, GA.

People

John Zhang

Optimization and Contact Simulation

Juan Alvarez Padilla

Learning and Optimization Based Controls

Sofia Kwok

Starfish Space

Taylor Howell

Google Deepmind

Zac Manchester

Assistant Professor

Last updated: 2025-03-07