Trajectory Optimization through Mixed-Integer Optimization of Contact Dynamics for Switching End Effector Locomotion
Jared Morgan, Mahdi Agheli
AI summary
Problem
Legged robots traditionally rely on fixed end-effectors, forcing a trade-off between adaptability and efficiency, while existing switching mechanisms lack optimal planning strategies for when and how to change contact dynamics.
Approach
The authors formulate a mixed-integer quadratically constrained control problem that jointly optimizes gait, contact timing, and end-effector selection based on terrain interaction, coupled with an extended whole-body controller for execution.
Key results
- First framework to jointly optimize gait and switching end-effector contact dynamics via mixed-integer programming
- Extended whole-body controller and state estimation to robustly execute switching trajectories in simulation
- Demonstrated reduced cost of transport and increased speed compared to foot-only locomotion
- Validated successful navigation across flat ground, ramps, and stepping stones on a simulated quadruped
Why it matters
Enables legged robots to dynamically exploit the efficiency of wheels and the adaptability of feet, advancing autonomous mobility in complex, unstructured environments.
Abstract
Trajectory optimizers for legged robots typically assume a single end effector on each leg, often a foot or wheel, without switching to another. Robots employing point-modeled end effectors, compared to those with wheeled end effectors, often benefit in adaptability and maneuverability but at the cost of higher energy expenditure and lower speed. While current hardware supports switching between these two end-effector types, existing research has largely focused on maintaining stability during switching, with little attention to determining when each type is most effective. To our knowledge, this paper introduces the first framework that simultaneously optimizes both trajectories and end-effector contact dynamics through mixed-integer optimization. We validate our approach by solv- ing and executing trajectories with a whole-body controller in Gazebo across a variety of terrains, including ramps and stepping stones. The results show that our framework not only handles diverse terrains but also exploits contact dynamics to reduce cost of transport and increase speed compared to foot- only locomotion.