Flexible intelligent microwave metasurface with shape-guided adaptive programming.

Empowering the reconfigurable metasurfaces (RM) with the capability to be mechanically deformable highlights the possibility to manipulate the electromagnetic (EM) wave across arbitrary surfaces. Such ambition is hampered by the absence of adaptivity to shape variation in current programming strategies for RM. Herein, we present a flexible intelligent surface platform (FISP) as a solution to achieve flexible RM with highly stable performance under dynamic deformation. The geometry acquisition module in FISP enables real-time awareness of RM's deformation with the conformal sensor array. By merging the actual shape of flexible RM into the input of the adaptive algorithm driven by the artificial neural network, the deformed flexible RM in FISP can be autonomously encoded by the bias voltage supply module to ensure robust performance under various deformation conditions. The versatility of FISP in manipulating EM waves is demonstrated by its applications in electromagnetic illusion, carpet cloaking, and data transmission, illustrating the prospects for seamlessly integrating flexible electronics and RM in the development of future EM metasurfaces.