Verizon and NEC Infuse AI Into Deployed Fiber - SDxCentral

Verizon and NEC used artificial intelligence (AI)-infused software and sensors as part of a recent proof-of-concept (PoC) trial that allowed already installed fiber infrastructure to monitor traffic and road conditions. The PoC used NEC-developed optical sensor technology with the AI software running alongside existing wavelength division multiplexing (WDM) communication channels on the same fiber lines. The mixture turned the cables into distributed optical sensors that collected information on traffic patterns, road conditions, road capacity, and vehicle classification information. The AI tools used included convolutional neural networks (CNN) and software vector machines that were able to tap into distributed intelligent traffic informatics (DITI). Those tools were plugged into a single integrated monitor that was able to detect back-scattering light traveling through the fiber optic streams to detect static strain, dynamic strain, acoustics, vibrations, and temperatures for each fiber segment.