Building the future of smart telecommunication systems with optical AI

Building the Future of Smart Telecommunication Systems with Optical AI 

Modern communication networks must handle ever‑growing volumes of data, driven by cloud services, connected devices, and real‑time applications. At the same time, they face a critical constraint: keeping energy consumption as low as possible. Today, signal recovery and data processing rely mostly on electronic hardware—powerful, but energy‑intensive and increasingly limited by latency. 

To address these challenges, researchers in Roberto Morandotti’s laboratory at the Institut national de la recherche scientifique (INRS) have developed a new device that enables optical artificial intelligence, where data is processed using light rather than electronics, enabling high speed and low energy consumption.  

This work, recently published in Nature Communications, demonstrates how photonics can offer a fast and efficient alternative to conventional signal processing technologies. 

The device belongs to the field of neuromorphic photonics, which takes inspiration from the way the human brain processes information.  

“Our research shows that optical AI can be a powerful alternative to traditional electronics,” says Luigi Di Lauro, INRS research associate and co‑corresponding author of the study. “By computing with light, our device can recover distorted telecommunication signals with high accuracy, while operating at ultra‑high speed and consuming only a fraction of the energy.” 

One Optical Device, Multiple Tasks 

At the core of the platform is a simplified artificial‑intelligence approach known as reservoir computing. This method allows complex machine‑learning tasks to be carried out using a compact and flexible architecture, where the same optical device can be reconfigured for different applications without changing any physical components. 

One of the technology’s key advantages is its ability to process multiple data streams simultaneously. By harnessing the natural speed and parallelism of photonics, the system enables real‑time signal processing with dramatically improved energy efficiency.  

“The intersection between photonics and AI opens a path toward faster, more energy-efficient communication networks.” 
— Professor Roberto Morandotti 

In the long term, integrating optical AI directly into existing telecommunications infrastructures could help build faster, more efficient, and more sustainable communication networks, supporting internet traffic, cloud services, and edge computing while reducing their environmental footprint. 

About the paper 

Aadhi, A., Di Lauro, L., Fischer, B. et al. Scalable photonic reservoir computing for parallel machine learning tasks. Nat Commun 17, 1225 (2026). https://doi.org/10.1038/s41467-025-67983-z  

This work was supported by NSERC through the Alliance and Discovery Programs, as well as by the Canada Research Chair Program, and Mitacs Elevate Program.