Design and Experimental Validation of a Bionic Robotic Capturing System Inspired by Net-Casting Spiders

Abstract

Capturing a noncooperative flying object is a challenging task that consists of two subtasks: precise tracking of the object and manipulation of the robotic grippers. High operation precision is required to avoid the object flying away from the gripper. In this article, we propose a bionic scheme to capture noncooperative flying objects imitating the predatory behavior of net-casting spiders. Based on this scheme, a robotic capturing system with a flexible net is developed for capturing noncooperative flying objects. The robotic capturing system consists of an unmanned aerial vehicle (UAV) platform, an optical tracking device, and a capture unit. These items imitate the body, median eye, and net of a net-casting spider. A scale adaptive tracking method is proposed to track noncooperative flying objects from video imagery collected by the optical tracking device used for guidance and control of the robotic capturing system. The capture unit deploys a net to entangle with the object. The proposed scheme is validated systematically to capture a noncooperative UAV and the experimental results demonstrated that the robotic capturing system has the advantages of robustness and adaptability.