Flexible tactile sensor based on machine learning inspired by bio mechanoreceptors

Mechanoreceptors in animals and plants play a crucial role in sensing mechanical stimuli such as touch, motion, stretch, and vibration. Learning from the mechanisms of mechanoreceptors may facilitate the development of bionic tactile sensors, leading to higher sensitivity, spatial resolution, and dynamic ranges. However, very little literature has comprehensively discussed the relevance of biological tactile sensing systems and machine-learning-based bionic tactile sensors. This review first introduces the structural features, signal acquisition and transmission mechanisms, and feedback processes of both plant and animal mechanoreceptors, and then summarizes the efforts to develop bionic tactile sensors by mimicking the morphologies and structures of mechanoreceptors in plants and animals. Additionally, the integration of artificial intelligence approaches with these sensors for data processing and analysis are demonstrated, followed by the perspectives on current challenges and future trends in bionic tactile sensors. This review addresses the challenges in developing high-performance tactile sensors by focusing on surface microstructures and biological mechanoreceptors, serving as a valuable reference for developing bionic tactile sensors with enhanced sensitivity and multimodal sensing capabilities. Furthermore, it may benefit the future development of smart sensing systems integrated with artificial intelligence for more precise object and texture recognition.