The future of artificial intelligence may depend not only on how powerful it becomes, but also on how much energy it consumes.
A study published in Nature Communications and coordinated by the Sant’Anna School of Advanced Studies in Pisa has developed an innovative bio-inspired tactile system capable of replicating how humans perceive and process touch while using significantly less energy.
The research was carried out in collaboration with STMicroelectronics, the University of Zurich and ETH Zurich, and represents an important step toward building robots and intelligent machines that can safely interact with their surroundings and with humans.
The system combines an artificial sensory “skin” equipped with optical sensors that imitate the behavior of tactile receptors in human skin with a bio-inspired neural architecture.
Signals generated by the sensors are processed by a neural network designed to replicate the structure and strategies of the human nervous system using a dedicated processor.
According to the researchers, this approach not only improves the precision of tactile signal processing compared with conventional AI systems, but also significantly reduces energy consumption.
The new technology could pave the way for a generation of highly efficient, low-power sensory systems for robotics and future AI applications, where sustainability and energy efficiency are becoming increasingly critical challenges.