Scientists have successfully developed an artificial tactile system that can be applied in prosthetic hand manufacturing, enabling robots to perceive sensations similar to humans.
The research was conducted by scientists at Uppsala University and the Karolinska Institute. Assistant Professor Libo Chen, the head of the research team, stated that they are working on developing a tactile system capable of sensing pain and temperature, as well as identifying the materials that the hand touches, such as distinguishing between wood and metal.
The artificial tactile system consists of three main components: electronic skin (e-skin) with embedded sensors that detect pressure through touch, an artificial neural network that converts sensory signals into electrical impulses, and a signal processor that analyzes and identifies objects. This system has the potential to learn to recognize an unlimited number of objects.
According to the researchers, humans will be able to interact with robots and prosthetic hands more safely and easily thanks to the tactile feedback. Moreover, the prosthetic hand will be able to manipulate objects as skillfully as a human hand.
“The electronic skin contains millions of sensory receptors. Currently, the e-skin cannot provide sufficient sensitivity, but new technology will make this feasible, so we aim to create artificial skin for the entire robot body,” Chen said.
The centrally oriented artificial tactile network used in hardware is made up of a series of sensors crafted from polyimide film shaped like a hand and a PCB circuit that converts signals into sequences. (Source: Science).
The tactile system can be applied in healthcare, for example, to monitor motor function disorders caused by Parkinson’s and Alzheimer’s diseases or to assist patients in rehabilitation after a stroke.
“Our technology can identify an object as quickly as a blindfolded person, simply by touching and feeling, it can indicate whether the object is a tennis ball or an apple,” Zhibin Zhang, a teaching assistant in the Department of Electrical Engineering at Uppsala University, stated.
During the experiments, Zhang and his colleague Libo Chen collaborated with researchers who specialize in data processing and machine learning from the Department of Systems and Signals (Uppsala University), along with a team from the Department of Neuroscience, Care Science, and the Department of Neurology (Karolinska Institute).
Drawing inspiration from neuroscience, the research team developed the artificial tactile system to mimic how the human nervous system responds to touch. This system uses electrical impulses to process information after each touch, similar to the human nervous system.
“This technology will help users feel the prosthetic hand as a natural part of their body,” Zhang explained. The research team tested the system using 22 different objects to grasp and 16 surfaces to sense.
22 objects were used in the grip test for object classification. (Source: Science)
According to Zhang, a member of the research team, the system could be further developed to predict when a patient is about to fall. The information could then be used to activate external devices to prevent the patient from falling or alert assistive devices to intervene.
