Whole-Body Humanoid MPC (wb_humanoid_mpc)

A whole-body nonlinear model predictive controller for humanoid loco-manipulation, built on an extended version of ocs2, with a Unitree G1 example.

Licenses

Code: BSD-3-Clause (opens in a new tab)
Read by Robotheca from LICENCE.txt (opens in a new tab) (BSD-3-Clause) at commit 6fb34bb (opens in a new tab), on 22 Oct 2026.
Commercial use: Yes

Optimizes humanoid motion in real time through full-order torque-level dynamics.
Offers a centroidal dynamics MPC that optimizes whole-body kinematics and center of mass dynamics, using either a single rigid body or full centroidal dynamics.
Offers a whole-body dynamics MPC that optimizes contact forces and joint accelerations, and can compute joint torques across the horizon.
Supports the Unitree G1 robot; a 1X Neo example is listed as coming soon.
Lets users command base velocity and root height via a robot base controller GUI or an Xbox joystick.
Provides Docker-based and local installation options built with colcon and ROS 2.

Written by Robotheca from the project's README (opens in a new tab) at commit 6fb34bb.

Difficulty: Advanced
Hardware it needs, as the project states it: Building consumes significant RAM: 16 GiB system RAM with PARALLEL_JOBS=2 (default), 32 GiB with 4, and 64 GiB with 6. Runs on ROS 2; Jazzy is named in the install guide.
How to start: The project's README (opens in a new tab)

Checked on GitHub on 22 Oct 2026.

No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.

Robotheca lists software as a reference: nothing is hosted or run here, and every fact links to where it was read.