부산대학교 도서관

To enhance the tactile perception and human–computer interaction efficiency of soft robots, an intelligent soft gripper system is proposed in this study. The system utilizes triboelectric nanogenerator sensors to capture tactile information during the grasping process. By employing a patterned electrode design, the sensor can extract multiple data, including the contact material, contact position, contact area, and slip. The conducted experiments revealed that the use of conductive gel instead of conventional metal electrodes improves compatibility with the soft gripper. The collected information is used to train the support vector machine (SVM) model, achieving accurate recognition of various objects with a recognition accuracy of 98.5%. Real-time communication analysis using the developed model demonstrates the recognition ability of the system when the software gripper interacts with objects, highlighting its application potential in human–computer interaction.