Design of a Mobile Manipulation Robot for Learning-Optimization Hybrid Grasping

Created to replace human workers, mobile manipulation robots exhibit substantial potential in addressing labor shortage problem shared by commercial and industrial businesses. A mobile manipulation robot combines the navigation ability of unmanned ground vehicle and manipulation advantage of robotic arm. However, the development of mobile manipulation robot is challenging due to integration of multiple heterogeneous subsystems. In this paper, we propose a mobile manipulation system by integrating perception, mapping, navigation, object detection, and grasping functions through a seamless workflow to conduct search-and-fetch tasks. Control Lyapunov Function (CLF) and Control Barrier Function (CBF) are employed to realize path tracking and guarantee safety during navigation. Convolutional Neural Network (CNN) and gripper's kinematic constraints construct the learning-optimization hybrid grasping algorithm to generate precise grasping poses. The efficiency of the mobile manipulation robot is demonstrated through indoor fetching experiments, showcasing its capabilities in real-world applications.