Triboelectric technologies for adaptive and self-powered neuromorphic tactile sensing

Abstract Tactile perception is fundamental to how humans interact with the world, underpinning both physical manipulation and emotional experiences. This complex sensory system relies on highly specialized mechanoreceptors embedded in the skin, which detect and relay signals through afferent neuronal fibers to the brain. Mechanoreceptors and afferent neurons perform local signal preprocessing, enabling transduction, filtering, adaptation, and sensory encoding before transmission to the central nervous system. While artificial tactile sensors inspired on the sophisticated functionalities of biological systems have been proposed, doing so in an energy efficient manner requires the integration of mechanical energy harvesters as transducers. Triboelectric nanogenerators have demonstrated significant potential as artificial mechanoreceptors, offering fully self-powered tactile sensing capabilities while simultaneously being able to supply energy to nearby low-power electronic components involved in perceptual processing. To bring significant progress in this technology, novel devices should emulate not only the different response of the various skin mechanoreceptors but also the first layer of tactile encoding processing accomplished at the skin level, while also minimizing energy consumption and environmental impact. Only few triboelectric (TE) effect driven tactile sensors have achieved the emulation of sophisticated adaptation features, such as slow and fast adaptation, typical of human tactile afferent neurons. Two approaches have been followed so far, which comprise the development of multifunctional TE-transducers capable of replicating more than one biological mechanoreceptor adaptive behavior, and fully integrated sensors embedding the TE-transducer and neuromorphic devices. Both approaches are still at their infancy, both in materials and device design, and still lack the high degree of integration, reproducibility and scalability that is required for their interface with standard electronics. Addressing such a pivotal challenge will improve neuroprosthesis integration, tactile robotics, human–machine interfaces, and the field of neuromorphic engineering for extreme edge processing.