Safety-critical Motion Planning for Collaborative Legged Loco-Manipulation over Discrete Terrain

Mohsen Sombolestan,Quan Nguyen

2024-10-15

Abstract:As legged robots are deployed in industrial and autonomous construction tasks requiring collaborative manipulation, they must handle object manipulation while maintaining stable locomotion. The challenge intensifies in real-world environments, where they should traverse discrete terrain, avoid obstacles, and coordinate with other robots for safe loco-manipulation. This work addresses safe motion planning for collaborative manipulation of an unknown payload on discrete terrain while avoiding obstacles. Our approach uses two sets of model predictive controllers (MPCs) as motion planners: a global MPC generates a safe trajectory for the team with obstacle avoidance, while decentralized MPCs for each robot ensure safe footholds on discrete terrain as they follow the global trajectory. A model reference adaptive whole-body controller (MRA-WBC) then tracks the desired path, compensating for model uncertainties from the unknown payload. We validated our method in simulation and hardware on a team of Unitree robots. The results demonstrate that our approach successfully guides the team through obstacle courses, requiring planar positioning and height adjustments, and all happening on discrete terrain such as stepping stones.

Robotics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the safe motion planning for collaborative legged locomotion and manipulation on discrete terrains. Specifically, the paper focuses on how multi - legged robots can handle objects while maintaining stable motion when they need to collaboratively carry objects in industrial and autonomous construction tasks. These tasks require robots to not only traverse discrete terrains and avoid obstacles in the real - world environment, but also coordinate with other robots to achieve safe collaborative carrying. The paper particularly emphasizes how to achieve this goal through two - level model predictive controllers (MPCs) - global MPC and local MPC, as well as a model reference adaptive whole - body controller (MRA - WBC) in the case of unknown loads. The global MPC is responsible for generating the safe trajectories for the entire team and considering obstacle avoidance; while the local MPC of each robot ensures the safe foot - position selection on discrete terrains while tracking the global trajectory. The MRA - WBC is used to track the desired path and compensate for the model uncertainties due to unknown loads. The paper verifies the effectiveness of the proposed method through simulation and hardware experiments, showing that the robot teams can successfully pass obstacle courses including planar positioning and height adjustment, and all this occurs on discrete terrains such as stepping stones.