Inside the Smart Tech That Predicts Volcanic Unrest

There have been between 60 to 70 volcanic eruptions in 2025. One of them occurred on June 2nd in Mount Etna. You may have seen footage of tourists running from the explosion. However, the Sicilian volcano is one of the most monitored in the world, to protect the residents living on its slopes and the 1.5 million visitors the fiery mountain attracts each year. If history has taught us anything, with Pompeii and more recently, La Palma, cities can be great victims of volcano eruptions, leading to infrastructure, nature, and human disaster. At Mount Etna, the National Institute of Geophysics and Volcanology, INGV, relies on a connected system to sense eruptions before they happen to stay ahead of danger. Basically, a volcanic eruption is a complex physical and chemical phenomenon that moves the magma from the crust up to the surface. That needs to be studied in different scientific disciplines. But here at the INGV, we don't rely on a single tool. We use, multi parametric approach, listening to the volcano in different ways. An eruption at Mount Etna is perceived by a series of signals, which INGV get thanks to a network of different types of sensors. There’s a large network of seismometers, sensors that listen for earthquakes and volcanic tremors, the vibration caused by magma underground. They also rely on infrasound sensors, for detecting low frequency sound waves that can come from explosions at the surface, which indicates the eruption is going to happen. They have other sensors, like tiltmeters and strainmeters that measure ground deformation, ground sensors that analyze CO2 and sulfur levels, and cameras to monitor the summit area. But they don’t just rely on on site sensors. The Institute also uses imagery data from satellites to see ground deformation. All the data is sent via radio, optic fiber, satellite connection, Starlink… It depends on where the sensors are located. Data is real-time and sent to control room servers. There, the Time Series Database System platform standardizes data. And this platform has been deployed here at the observatory in Catania by my work team and it’s called TSD system. That stands for Time Series Database system. In a few words, this is a system that allows the operators and researchers to have a complete and constantly updated view of the volcano's behavior. Different softwares run on the platform to then display the information in the control room, where the INGV team monitors 24/7 the data that is being processed. There, the team sees the data in a visual manner, data like tremor alerts, or the camera feed. AI also plays a role in the identification of issues, as it analyzes images from satellites. The Institute is also developing an AI tool, dubbed ETNAS, to analyze volcano information and give early warnings to civil protection agencies of events of lava fountains. ETNAS is set to send automatic alerts to agencies as soon as it identifies a possibility, giving the warning first and then allowing the team at the control room to explore how important the event is set to be. A key solution considering how important time is if evacuation is needed. We are particularly lucky because Etna is quite predictable in this sense. So the performance are quite good. We have 99% of probability to detect correctly a paroxysm, with one hour and a half, on average, of lead time. The monitoring of the volcano is leading to a safer area. And it also has a clear ROI. It's all about keeping people safe and sending alerts early. But it also helps Sicily keep up one of its main incomes. Tourism. Before there was smart tech, anytime an event occurred, Mount Etna and its surroundings would be closed off for long periods of time, as they couldn’t predict what was happening. Now, the park knows exactly when to close and when the danger has passed, keeping tourism levels high, which can in turn help them to keep investing in tech on Mount Etna.