<p>Skin-conformal electronics that naturally adhere to the body are transforming human–machine interfaces (HMIs) by enabling intuitive, real-time gesture recognition with broad potential in immersive applications such as virtual reality, advanced robotics and remote health care. These devices bridge human intentions and machine responses but still require integrated platforms that collect advances in sensing elements, adaptive signal processing and intelligent decision-making algorithms. In this Review, we identify the component innovations and system-level strategies to define a roadmap for skin-conformal gesture recognition as a core element of next-generation HMIs. Advances in conformal device architectures are overcoming the mechanical and signal stability limitations of conventional wearables, enabling reliable operation during continuous use. Integration of sensing and processing enables adaptive, real-time interpretation of gestures aligned with user intent, while emerging computational approaches deliver efficient, low-latency performance inspired by biological learning. Collectively, these developments are shaping design principles for natural, precise and responsive HMIs.</p>