Intrinsically Linear and Robust Capacitive Strain Sensors Based on Area‐Variation Mechanism for Soft Machines

Flexible capacitive strain sensors are essential for soft machines, yet conventional distance‐variation designs inherently exhibit nonlinear responses and limited mechanical robustness. Here we report an intrinsically linear capacitive strain sensor based on an area‐variation mechanism. The sensor employs laser‐micromachined carbon fiber interdigitated electrodes fully encapsulated in polydimethylsiloxane. This yields a highly linear electromechanical response, while experimental deviations in baseline capacitance are attributed to parasitic wire capacitance. For a typical sensor, the electromechanical response exhibits exceptional linearity ( R 2 = 0.9995) and a high sensing accuracy of 0.381% full scale across the entire strain range. The sensor preferentially responds to axial tension with small response of normal compression, while retaining stable performance following over‐range stretching, mechanical impact, and ten water immersion‐drying cycles. When integrated into a smart glove, the sensor enables real‐time, high‐linearity capture of finger‐bending strains for accurate gesture recognition. Embedding the sensor in a cable‐driven soft continuum robot further allows precise monitoring of microstrains in actuation wires, supporting real‐time shape reconstruction with accuracy comparable to that of commercial force sensors. This area‐variation sensing strategy resolves the long‐standing tradeoff between linearity and sensitivity in flexible capacitive strain sensors, offering a robust and scalable solution for advanced wearable systems and minimally invasive surgical robots.