Temperature–strain decoupled electronic skin for human–machine interaction

Electronic skins are promising for wearable motion monitoring and human–machine interaction owing to their mechanical compliance and conformal contact with human skin. However, in practical environments, piezoresistive strain sensors are intrinsically susceptible to temperature fluctuations, where thermally induced resistance variations are superimposed on strain signals, leading to baseline drift and sensing inaccuracies under practical conditions. Herein, we report a flexible electronic skin platform with real‐time temperature–strain decoupling capabilities. A serpentine strain sensor and a miniaturized thermistor are co‐integrated within a compliant multilayer architecture to simultaneously monitor mechanical deformation and local temperature fluctuations. Through multi‐temperature and multi‐strain calibration, a two‐dimensional regression model is established to quantitatively distinguish the thermal and mechanical contributions to the resistance response. The extracted effective temperature coefficient is further embedded into the acquisition algorithm, enabling real‐time subtraction of temperature‐induced drift. The device exhibits stable temperature and strain responses, fast response/recovery times of 100/200 ms, and reliable performance over 10 000 loading‐unloading cycles. The integrated thermistor maintains strain‐insensitive temperature sensing performance, providing a robust thermal reference under dynamic thermal conditions. The decoupled output effectively suppresses thermally induced baseline drift while preserving the strain‐induced waveform. In human motion monitoring, the E‐skin platform enables stable sensing during bending, stretching, swallowing, and torsional deformations. The decoupled signals are further wirelessly mapped to a four‐degree‐of‐freedom robotic arm, realizing real‐time human–machine interaction. This work provides a physically interpretable and potentially scalable strategy for temperature‐resilient wearable sensing systems operating in dynamic environments.