Cooperative Transportation With Mobile Manipulator: A Capability Map-Based Framework for Physical Human–Robot Collaboration
Heng Zhang, Sheng Qi, Jiawei Hu, Xinjun Sheng, Zhenhua Xiong, Xiangyang Zhu
Abstract
In the cooperative transportation task with a mobile manipulator (MM), the mobile robot and the manipulator must move simultaneously to adapt to the human motion. In addition, the motion of the MM is underconstrained due to redundancy, which makes MM real-time motion planning challenging. In this article, a capability map-based framework was proposed to enable real-time motion planning of the MM. In the motion planner, the dexterity of the MM, the formation of the human–robot system, and obstacle avoidance of the mobile robot are considered to get safe and human-like robot motion. Moreover, to make optimization faster and avoid local minimum, capability map, which represents the manipulability distribution of the MM in its workspace, is queried to determine the seed of the optimization algorithm. The proposed motion planner is self-contained and can be extended to the transportation task with multiple MMs. The effectiveness of this proposed framework is validated both in simulations and real-world experiments.