Artificial skin that “feels” like human skin: a study coordinated by the Sant'Anna School of Advanced Studies in Pisa opens up new prospects in bionics and collaborative robotics
A research team has developed an artificial skin that localises touch with human-like precision, thanks to a bio-inspired neuronal network that replicates the sensory mechanisms of the nervous system. This innovation, based on fundamental discoveries related to the somatosensory system, marks a significant advance for smart wearable technologies, prosthetic devices that provide realistic sensory feedback, and anthropomorphic robots
A research team from the Scuola Superiore Sant'Anna in Pisa, coordinated by Calogero Oddo, associate professor of Bioengineering, has developed an artificial skin that emulates the morphological characteristics and tactile functions of human skin thanks to innovative sensors and artificial intelligence algorithms inspired by the neural structures that convey and process tactile information.
The study, the result of a collaboration between the Neuro-Robotic Touch Lab of the Biorobotics Institute of the Sant'Anna School of Advanced Studies and the Federal University of Uberlandia (Brazil), was published in the international journal Nature Machine Intelligence: In addition to offering a fundamental contribution to the understanding of the mechanisms of human touch, the research opens up new perspectives in the fields of bionics and collaborative robotics for the development of intelligent wearable devices, prostheses capable of providing information on tactile interaction and sensors that allow robots to perceive their surroundings.
"The perception of touch is an essential function for human beings: it allows us to recognise and locate physical stimuli, explore the environment and interact safely with the external world. Artificially reproducing this complex sense is one of the main challenges in the design of collaborative robots and bionic prostheses" says Calogero Oddo.
The biomimetics of sensorized skin
The technology developed by the Sant'Anna School not only reproduces the sensitivity of human skin, but also replicates the logic by which the brain interprets and localises tactile stimuli. At the heart of the innovation is a large-area artificial skin equipped with fibre optic sensors capable of detecting pressure and touch in real time. This skin is made “intelligent” by a bio-inspired computational architecture for tactile information processing, a network of spiking neurons designed to mimic the mechanisms of the human nervous system.
"The spiking neuronal network consists of two layers: the first simulates human type 2 mechanoreceptors (with slow and fast adaptation), while the second reproduces a somatosomatic map similar to that generated by the neurons of the cuneate nucleus, a key region in tactile perception. This approach allows the artificial skin to identify the point of contact and decode the intensity of the stimulus" says Mariangela Filosa, assistant professor at the BioRobotics Institute of the Sant'Anna School of Advanced Studies and first co-author of the study.
Future applications and scenarios
The potential applications of this technology are manifold: from collaborative robotics, to ensure safe interaction between people, environments and machines, to bionic prostheses, with the development of new sensory feedback solutions for people with disabilities.
"Since the fourth industrial revolution, interaction between humans and machines has become a key element in many areas of robotics. The possibility of equipping robots with an artificial sense of touch inspired by the human sense makes this interaction safe, intuitive and effective, mitigating the risk of accidents for operators. In addition, the artificial skin we have developed also offers broad prospects in the field of bionics: cutting-edge assistive and rehabilitative technologies will be equipped with intelligent tactile sensors to restore or enhance tactile information through realistic, cutaneous or neural feedback" comments Calogero Oddo.
