New sensor gives robots lifelike touch and instant grip control

Researchers at the University at Buffalo have developed a novel fabric-based electronic textile (E-textile) sensor that mimics human skin’s ability to sense pressure, slippage, and movement. This sensor, integrated onto 3D-printed robotic fingers and a compliant gripper, enables robots to react in real time by adjusting their grip strength dynamically—tightening or loosening as needed. The sensor operates via the tribovoltaic effect, generating electrical signals from friction, and demonstrates response times as fast as 0.76 milliseconds, surpassing typical human touch receptor speeds. This rapid and sensitive feedback allows robots to detect object slippage and adjust their grasp without crushing items, significantly enhancing robotic dexterity. The breakthrough holds promise for a variety of applications requiring fine motor control and delicate touch, such as collaborative manufacturing tasks, robotic surgery, and prosthetic limbs. The research team plans to incorporate reinforcement learning to further improve control algorithms and explore the sensor’s adaptability across different robotic platforms. Funded by the University