An open-source parallel gripper with an embedded soft skin fingertip sensor

摘要

The demand for implementing robots into our daily lives has surged in recent years, necessitating safe grasping for effective interaction with the environment. However, a majority of researchers rely on commercial grippers for their experimental studies, which are typically expensive and not accessible to everyone. Despite the existence of open-source designs, the assembly process is often challenging and requires modifications to enhance secure grasping. This paper presents a simple, compact, and low-cost gripper to offer an accessible and readily deployable solution for research and education. The gripper utilizes a parallel four-bar linkage mechanism, minimizing the number of components and incorporating off-the-shelf parts for straightforward assembly. Furthermore, to enhance its capabilities, the proposed gripper implements a soft skin tactile sensor on its fingertips. These sensors offer three-directional measurements using Hall effect sensing and embedded silicone. By controlling fingertip force based on information from the tactile sensors, the gripper achieves safe grasping. The gripper is evaluated to grasp daily life objects with different properties such as shapes, sizes, and levels of deformability. Evaluation results showcase the gripper's versatility, enabling it to securely grasp various objects, including fragile items. This outcome underscores the gripper's effectiveness, versatility, and safety in practical use.