Stable sensing on a complex, moving surface depends first on the platform: it must stretch with the surface and stay conformally attached, while the circuit it carries keeps functioning. We have developed a bioinspired root-like interface that anchors rigid components into the stretchable substrate and delays interfacial failure [1], a thermoforming process that turns predistorted planar circuits into conformal three-dimensional electronics [2], a skin-conformal film whose enhanced adhesion and micro-wrinkled surface maintain contact during joint motion [3], and an auxetic-reinforced substrate with a negative Poisson’s ratio for omnidirectional stretching [4]. Together these provide the mechanical basis for electronics that operate on a moving surface.

Once the platform holds, the sensing layer must read the physical stimuli acting on that surface and tell them apart. Machine learning increasingly helps with this, from inferring a specific value to separating signals that overlap. We have developed a thermoformed three-dimensional film architecture for customizable tactile sensing [1], a strain sensor whose aligned carbon nanotube network makes its optical response direction-dependent [2], a battery-free multi-modal array that combines pressure, shear, and temperature sensing [3], and a liquid-metal microchannel pressure sensor that resolves force direction through machine learning [4]. These sensors supply the physical information needed for health monitoring, motion tracking, and interactive systems.

An integrated wearable system operates in two directions. Mounted on the body, a wearable sensor gives quantitative form to the signals arising on and around it, turning them into electrical output that can drive the electronics carried alongside it. That output can also travel onward into a virtual environment, giving the wearer an extended and immersive experience: a wireless, battery-free multi-axial sensor streams skin-interface data to augmented reality glasses [1], and a flexible-island wristband works as a stretchable controller that also returns haptic feedback [2]. Mounted on a machine, the sensor lets a robot read the stimuli around it as a person would, so that touch becomes a means of interacting with the people and objects nearby; in our work this takes the form of a modular tactile platform that covers a robot arm and registers contact with someone working beside it [3]. Both directions meet at the interface between people and machines.