The University of Cádiz has awarded the atrÉBT! prize in the Projects category to Deep Noise, the distributed acoustic intelligence initiative developed by Algeciras native Víctor Manuel Díaz, founder of SonaGrid. The award was presented in the courtyard of the Faculty of Economic and Business Sciences in Cádiz, during the gala of the XVIII edition of a call that this year distributed thirteen distinctions among 119 applications submitted by the university community and its environment.
The official project description summarizes its field of application: Deep Noise "develops networks of distributed acoustic intelligence for smart cities and critical infrastructures, capable of identifying noise sources in real time with privacy guarantees." The technology is based on a premise different from that of conventional sound monitoring systems, which only measure decibels at fixed points and compare the result with regulatory thresholds.
SonaGrid, the company supporting Deep Noise, is defined as a deep technology company specialized in the digitization and analysis of industrial and urban soundscapes. According to its profile on the innovation portal of the Port Authority of the Bay of Algeciras (APBA), its proposal involves transforming the management of environmental noise and replacing traditional decibel measurement with the generation of operational data through artificial intelligence. The company designs and manufactures its own autonomous IoT nodes, with energy independence and long-range connectivity to operate in critical environments.
The differential component lies in the processing. The algorithms of edge computing run on the node itself, without needing to send the audio to a central server, allowing for real-time characterization of the spectral footprint of each sound and identification of the specific source emitting it. The architecture, according to the same profile, offers infrastructure managers and public administrations a continuous monitoring tool to incorporate the sound dimension into daily operations.
