Nano-technology and Artificial Intelligence for Next-gen Optical information Storage

Classical optical storage concepts like the Blu-ray are physically constrained in their information density by the optical diffraction limit. We propose a disruptive concept for next generation optical storage, combining silicon nanostructures, photonics and machine learning (ML). We design nano-structures such that each geometry encodes several bits of information, each geometry is optimized to yield a unique optical scattering spectrum. Spectral optical measurements in combination with a ML-based data interpreter then robustly retrieve digital information denser than the diffraction limit. This project has as goal to maximize the number of bits per nanostructure, to develop advanced 3D fabrication techniques and invent optical read-out schemes to read data from stacked layers of nanostructures. It bears the potential to pave the way to true next-gen storage technology, giving at a time an answer to the exploding demand for data capacity and to its energy consumption.