Mechanics-guided design of flexible pressure sensors: bridging multiscale electromechanical mechanism to tactile-enabled embodied intelligence

As artificial intelligence (AI) transitions from the virtual to the physical realm, embodied intelligence equips agents with real-time interaction and feedback capabilities. Within the closed-loop ‘perception-decision-execution’ paradigm, environmental pressure feedback governs action decisions, while concurrent actions dynamically reshape perception. Consequently, adaptive and precise perception serves as the cornerstone of embodied intelligence. Flexible pressure sensors are pivotal in enabling tactile perceptions, translating subtle mechanical stimuli into measurable electrical signals via intrinsic electromechanical coupling mechanisms. This functionality allows intelligent systems to discern physical properties such as texture, shape and composition. This review systematically summarizes the mechanics-guided design of flexible pressure sensors, bridging multiscale electromechanical coupling behaviors with diverse tactile perception applications. It begins by elucidating the fundamental transduction mechanisms, including capacitive, piezoresistive, piezoelectric, and triboelectric, alongside theoretical advances in regulating electromechanical behaviors. Subsequently, it discusses fabrication techniques, emphasizing how microstructural design influences sensitivity, stability, and dynamic responses. Finally, it addresses the integration of high-performance pressure sensors with AI for tactile-enabled embodied intelligence, advanced perception and adaptive control. Looking ahead, interdisciplinary collaboration across mechanics, material science, and computer science is essential to unravel persistent challenges in electromechanical coupling and propel the development of multifunctional tactile sensory systems. © 2026 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.