Reproduce “Multi-Contact Motion Retargeting using Whole-body Optimization of Full Kinematics and Sequential Force Equilibrium” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Quentin Rouxel, Kai Yuan, Ruoshi Wen, Zhibin Li
Researchers from the University of Edinburgh developed SEIKO (Sequential Equilibrium and Inverse Kinematics Optimization), a framework for real-time teleoperation of high-DoF robots in multi-contact scenarios. SEIKO translates human operator commands into physically feasible and stable whole-body motions by optimizing full kinematics and sequential force equilibrium, demonstrating an average control cycle time of 0.81 ms on an embedded mini-PC.
Key insights
- Optimizing incremental changes in robot state rather than absolute configurations guarantees a valid 'no change' solution, critical for safety in teleoperation and ensuring stable halts.
- Combining full kinematics with sequential force equilibrium within a single optimization framework allows for dynamic adaptation of both robot posture and contact forces to maintain stability.
- Real-time performance at 1 kHz is achieved by performing only one SQP iteration per control loop, leveraging analytical partial derivatives and decomposing the equations of motion for efficiency.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.