Deep learning-enabled diffractive processors for low-power structural vibration monitoring

Structural Health Monitoring (SHM) is required to maintain the safety and longevity of civil infrastructure. Here, we present a diffractive vibration monitoring system, integrating a jointly-optimized diffractive optical processor with a shallow neural network-based backend to remotely extract 3D structural vibration spectra, offering a low-power and cost-effective solution. This architecture eliminates the need for dense sensor-arrays or extensive data acquisition. It instead uses a spatially-optimized diffractive optical processor that encodes 3D structural displacements into modulated light, captured by a minimal number of detectors and decoded in real-time by shallow and low-power neural networks to reconstruct the 3D displacement spectra of structures. The diffractive system's efficacy was demonstrated both numerically and experimentally using millimeter-wave illumination on a laboratory-scale building model with a programmable shake table. Our system achieves more than an order-of-magnitude improvement in accuracy over conventional optics or separately trained modules, establishing a foundation for high-throughput 3D monitoring of structures.