Force distribution of thumb-index finger power-grasp during stable fruit grasp control

To investigate the force distribution of the thumb-index finger power-grasp during the stable grasping of fruit, four thumb-index finger power-grasping experiments were performed. Fruits implanted with force-sensing resistor sensors with seven channels were used to test the real-time normal force of each thumb-index finger region. There was a significant difference among power-grasp regions, fruit size, between power-grasp postures, and centre of mass for the percentage contribution of normal forces on some regions (P < 0.05) during the power grasping of fruit using the thumb-index finger. The percentage contribution of normal forces on the distal phalanx of the thumb and index finger was larger than those on other regions, and gradually decreased from the distal phalanx to the proximal phalanx. There was a very weak linear association between the characteristics of the thumb and index finger and the percentage contribution of normal forces on each region. The longitudinal power-grasp is a more labour-saving posture than the horizontal power-grasp when grasping a fruit using the thumb and index finger. The normal force contribution on the distal and proximal phalanx of the index finger and the three combination regions are sensitive to changes in the centre of mass location. The normal force of the thumb and the tangential force should be increased to overcome the positive torque when the centre of mass is on the left side. Conversely, when the centre of mass of the fruit is located on the right side, the normal force of the index finger should be increased to prevent it from rotating and sliding to ensure the stable power-grasp of the fruit. It's necessary to adjust the normal force distribution on each grasping region of the humanoid end-effector based on the centre of mass location. This study reveals the force distribution mechanism considering the different characteristics of the fruit and hand and provides guidance for stable bionic two-finger robot grasp control.