UC Berkeley

Robotic grasping and manipulation are essential for applications ranging from industrial automation to assistive technologies. In this dissertation, I address the challenges of improving these capabilities through advancements in gripper design by leveraging intrinsic functionality in the systems. The first focus is on developing a smart suction cup system with pressure-based tactile sensors that detect contact status, enabling real-time adjustments for enhanced grasp success rates. A developed suction cup and its haptic search algorithm leverage suction airflow into tactile feedback to achieve robust robotic grasping. The second focus is on regaining grasping function through the development of an assistive Dorsal Grasper designed for individuals with spinal cord injuries at cervical levels. This device enhances grasping capability by leveraging residual motor functions and providing additional mechanical support, thus improving the ability to perform daily tasks. By leveraging the inherent functions of systems, the dissertation offers a simple but effective approach to advancing robotic and human-robot systems, contributing novel insights and solutions.