Real-Time Seismic Monitoring Using DAS and AI

Seismic monitoring in active volcanic areas requires systems capable of providing high spatial and temporal resolution, in order to capture the complex and rapidly evolving dynamics of such environments. In this context, the reuse of existing telecommunications infrastructure through distributed acoustic sensing (DAS) technology represents a promising solution, as it transforms standard optical fibres into dense arrays of virtual sensors distributed over large distances. This approach enables continuous, high-sensitivity monitoring and significantly improves the detection capability of local seismic activity, even for low-magnitude events.

In this study, we present an integrated seismic monitoring system for the Campi Flegrei area based on the combined use of artificial intelligence (AI) and DAS technology. The system has been implemented along the 22 km Bagnoli–Bacoli route by leveraging an existing optical fibre link. The continuous data stream acquired by the DAS system is processed in real-time by a cascade of AI models, specifically designed for the automatic detection of seismic phases (P and S waves), event association and localization, as well as magnitude estimation. This multi-stage architecture allows robust and efficient analysis of large volumes of data, enabling near real-time detection and characterization of local seismicity.

The processing results are stored in a structured database and made accessible through web services, including real-time visualization interfaces for monitoring and analysis. This architecture demonstrates how the integration of existing telecommunications infrastructure, distributed sensing technologies, and advanced AI methods can provide an effective, scalable, and low-cost solution for continuous seismic monitoring in complex volcanic areas. Such systems have the potential to enhance early warning capabilities and support risk mitigation strategies in densely populated regions exposed to volcanic hazards.