Scientists at the University of Buffalo have unveiled a groundbreaking electronic textile (e-textile) that empowers machines to detect objects by replicating the sensory capabilities of human hands. This innovative technology mimics how nerves in the human body perceive pressure and slippage when grasping items.
Jun Liu, PhD, an assistant professor in the UB Department of Mechanical and Aerospace Engineering and a core faculty member of the RENEW Institute, leads the study as the corresponding author. He emphasizes the potential applications of this technology in various fields. “This advancement could transform manufacturing processes, enhancing tasks like product assembly and packaging—essentially any scenario where humans and robots collaborate,” Liu explained. He also highlighted its importance in improving robotic surgical instruments and prosthetic limbs.
According to Vashin Gautham, a PhD candidate and the study’s first author, “Our sensor operates similarly to human skin. It is flexible, highly sensitive, and adept at detecting not only pressure but also subtle movements and slippage of objects.” This innovation aims to revolutionize the interaction between robots, prosthetics, and human-machine systems.
Researchers have successfully integrated this sensing technology onto a pair of 3D-printed robotic fingers, which are attached to a compliant robotic gripper designed by Ehsan Esfahani, PhD, an associate professor in the same department. “With this sensor, the robotic gripper can detect slippage and adjust its grip force dynamically, facilitating more complex in-hand manipulation tasks,” Esfahani noted.
For instance, when the team attempted to pull a copper weight from the robotic fingers, the gripper immediately tightened its grip in response to the sensed movement. “This sensor is the crucial element that brings robotic hands closer to mimicking human functionality,” Esfahani added. The slight movement between the object and the grip generates direct-current (DC) electricity due to the tribovoltaic effect, enhancing the sensor’s responsiveness.
The researchers have measured the sensor’s response times and found them comparable to human touch. Depending on the specific test, the system responded within 0.76 to 38 milliseconds, aligning with human touch receptors, which typically respond between 1 and 50 milliseconds. “The system is remarkably fast and meets biological benchmarks for human performance,” Liu commented. “Interestingly, a stronger or faster slip results in a more robust sensor response, which simplifies the development of control algorithms for precise robotic actions.”
Looking ahead, the research team plans to conduct further testing of the sensing system, including the integration of reinforcement learning—an artificial intelligence technique that could enhance robotic dexterity significantly. This project received support from the University at Buffalo Centre of Excellence in Materials Informatics, paving the way for potential applications that could redefine human-robot interactions.
