Key points are not available for this paper at this time.
Abstract A tactile sensor system enables natural interaction between humans and machines; this interaction is crucial for dexterous robotic hands, interactive entertainment, and other smart scenarios. However, the lack of sliding friction detection significantly limits the accuracy and scope of interactions due to the absence of sophisticated information, such as slippage, material and roughness of held objects. Here, inspired by the stick-slip phenomena in the sliding process, we have developed a multifunctional biomimetic tactile system based on the stick-slip sensing strategy, which is a universal method to detect slippage and estimate the surface properties of objects by sliding. This system consists of a flexible fingertip-inspired tactile sensor, a read-out circuit and a machine-learning module. Based on the stick-slip sensing strategy, our system was endowed with high recognition rates for slippage detection (100.0%), material classification (93.3%) and roughness discrimination (92.8%). Moreover, robotic hand manipulation, interactive games and object classification are demonstrated with this multifunctional system for comprehensive and promising human–machine interactions.
Li et al. (Fri,) studied this question.