Dexterous robotic grasping of delicate fruits aided with a multi-sensory e-glove and manual grasping analysis for damage-free manipulation

Gripping control of the manipulator is still a technical problem in its research field. Accurate, efficient, and lossless crawling remains a challenge due to the difficulty of controlling grip mode, joint movement, and optimal implementation. Human hands can accurately perceive objects and grasp flexibly, so understanding and simulating the gripping behavior of human hands is crucial for controlling manipulators. In this study, an electronic glove was designed and assembled based on manual grasping and mechanical grip perception, combined with force sensors, bending sensors, and inertial sensor (Inertial Measurement Unit, IMU). The mechanism and characteristics of handgrip are analyzed from the four aspects, including force and bending relationship, hand grasping patterns, finger cooperation and correlation, and the law of external interaction acceleration. The pressure generally increases with the increase of bending, so the force can be controlled by controlling the degree of bending. Moreover, mode A shows a more complex grip mechanism compared with mode B. The pose position and orientation of the manipulator can be controlled by analyzing the gripping characteristics between the various modes. Besides, the correlation between the index finger, middle finger, and ring finger is relatively high. By understanding the correlation law between the fingers can control the cooperative movement of the manipulator when grasping. And the angular velocity of the X and Y axes changes significantly during the grasping process. The direction of movement and position of the manipulator can be determined by the changing state of the angular velocity. Thus, the analysis of these aspects could provide theoretical guidance for the structural design, surface material selection, and grasping planning of bionic manipulators.