Tactile-sensing-based robotic grasping stability analysis

Abstract

Tactile signals play a crucial role in enabling robots to successfully manipulate unfamiliar objects. For robots to grasp unknown objects securely and without causing damage, it is essential that they can analyze grasping stability in real time through tactile signals and respond promptly. This study introduces a novel method for analyzing the stability of robotic hand grasping, utilizing the Wilcoxon signed rank test. The efficacy of this method is demonstrated through its static and dynamic performance, and evaluated across a series of experiments. The findings of this research highlight the method’s ability to accurately detect when an object begins to slip from the robot’s grasp. Employing this method allows the gripper to maintain a secure hold on objects by applying the minimal necessary force. It also enables the gripper to dynamically adjust the force it applies in real time, thus preventing the object from slipping during the movement of the robotic arm. Moreover, the gripper demonstrates the ability to stably grasp objects of varied weights and with unknown characteristics, showcasing the versatility and effectiveness of the proposed method.