We are pleased to announce version 1.0 of pymanoid!
pymanoid is a Python controller prototyping environment for humanoid robotics
based on OpenRAVE. Its main features are:
• Whole-body inverse kinematics (IK) based on the weighted multi-task framework
• Contact-stability conditions: multi-contact ZMP areas, wrench cones, etc.
• Linear and Nonlinear Model Predictive Control (MPC) for locomotion
• Interfaces to geometry and numerical optimization (LP, QP and NLP) solvers
The following use cases are provided (all open-source):
• Walking pattern generation over uneven terrains with variable CoM height
• Nonlinear predictive control by direct transcription of centroidal dynamics
• Multi-contact ZMP support areas for locomotion including hand contacts
• Humanoid stair climbing demonstrated on the HRP-4 humanoid robot
The project is fully open-source. For examples, tutorials and more, see: