Researchers at the University of California, Los Angeles (UCLA), have unveiled a groundbreaking framework for Structural Health Monitoring (SHM) that leverages artificial intelligence and diffractive optical processors. Led by Professor Aydogan Ozcan, the team has developed a system capable of monitoring structural vibrations with unprecedented energy efficiency and simplicity, offering a powerful alternative to traditional sensor networks. Current SHM systems are essential for assessing the integrity of critical infrastructure, such as bridges and buildings, especially following natural disasters. However, conventional methods rely heavily on networks of accelerometers and strain gauges. These electronic sensors are often energy-intensive, generate massive amounts of data that require complex digital signal processing, and are costly to install and maintain. Furthermore, achieving the high spatial resolution needed to accurately pinpoint damage typically demands a dense array of sensors, further inflating costs. The new UCLA technology addresses these limitations through a unique physical-digital co-integration approach. Instead of digitizing raw physical signals at the source, the system utilizes a passive, AI-optimized diffractive layer attached directly to the structure. As the building or bridge vibrates, this specialized surface moves, modulating an incoming light wave to encode structural displacements directly into optical patterns. This process transforms complex mechanical oscillations into distinct spatiotemporal light signals. Unlike traditional setups, the diffractive layer acts as an intelligent optical processor that pre-encodes multidimensional information physically. Only the resulting modulated light signals need to be captured by a minimal number of optical detectors. These signals are then rapidly decoded by a low-power shallow neural network. This design shifts a significant portion of the computational burden from the digital domain to the physical domain, drastically reducing power consumption and data processing requirements. In experimental validation, the researchers collaborated with experts from UCLA's Civil and Environmental Engineering Department and the California Geological Survey. Using a laboratory-scale building model on a programmable shake table, they successfully illuminated the diffractive surface with millimeter waves. The system accurately extracted one-dimensional and two-dimensional vibration spectra under various dynamic conditions, including seismic waveforms from actual earthquake datasets. Additionally, the team demonstrated a wavelength-multiplexed version of the system, capable of monitoring multiple vibration points simultaneously using different light wavelengths. A key advantage of this innovation is its scalability and energy efficiency. The diffractive surface is entirely passive, consuming zero energy during its encoding operation. Moreover, the design is adaptable across the electromagnetic spectrum. By adjusting the dimensions of the diffractive features relative to the illumination wavelength, the technology can be scaled to operate using visible or infrared light, not just millimeter waves. This flexibility suggests broad applicability for future infrastructure monitoring, paving the way for low-power, high-resolution health assessment systems that are cost-effective and easy to deploy.