Galicia Deploys AI Camera Network to Detect Wildfires Ahead of Megafire Season

Authorities in Galicia, in northwestern Spain, launched an artificial intelligence system ahead of this year's fire season aimed at cutting wildfire detection time from hours to minutes. The system, based on convolutional neural networks, analyzes real-time footage from 241 cameras placed in the region's highest-risk parishes.

Response to a Record Season

The decision came after the worst fire season in the region's history. In 2025, fires consumed more than 118,000 hectares of forest and farmland in Galicia, an area larger than the entire city of Berlin. The fire in Larouco, in the province of Ourense, became the largest single wildfire in the region's recorded history. The scale of destruction has led authorities to speak of a megafire phenomenon, which, driven by rising temperatures and drought, is changing in character and intensity compared with fires from a decade ago.

The European Union as a whole experienced its most destructive wildfire season on record in 2025, with more than a million hectares burned across 25 of its 27 member states. Along with Portugal and Greece, Galicia has for years been among the European regions most exposed to wildfires, due in part to dense forests, an aging rural population, and abandoned farmland that easily becomes overgrown with highly flammable vegetation.

How the System Works

The new system has been built into Xeocode, the internal fire management platform used by Galicia's regional emergency services. The AI models were trained to recognize the characteristic rising columns of smoke, heat signatures, and specific color patterns in the landscape that distinguish a wildfire from other phenomena, such as fog or dust.

According to data from a two-year pilot testing period, the system can detect a fire from more than 15 kilometers away, and up to 25 kilometers under favorable weather conditions. When the algorithm flags a potential fire, the alert first goes to a technician who verifies the signal before firefighting crews are dispatched to the scene. This extra verification step is meant to reduce false alarms while keeping response times fast.

Investment Plan and Regional Comparisons

The camera network is part of a broader plan that Galician authorities describe as the largest wildfire prevention program in the region's history. Besides the AI technology, the 213 million euro investment includes a new app for residents and emergency services, as well as additional personnel and equipment. The system came online even before July 1, 2026, the official start of the region's high-risk fire season.

Galicia is not the only region in southern Europe turning to artificial intelligence to fight wildfires this year. Greece became the first country to add a dedicated satellite constellation to its national system: four nanosatellites built by German company OroraTech that scan the country's territory twice a day. The satellites' thermal sensors can detect fires as small as four meters across, with AI-processed data going directly to national fire brigade commanders. The models also filter out false alarms triggered by things like heated rocks, solar panels, or metal factory roofs.

Portugal, the country with the largest cumulative burned area in Europe over the past two decades, is also incorporating satellite detection into its national procedures, combining ground cameras, satellite data, and machine learning into a single early warning system.

The changes are meant to adapt the system to new types of fires of enormous intensity - Alfonso Rueda, president of the Xunta de Galicia

What It Means for Poland

Although the scale of wildfire risk in Poland is far smaller than in Mediterranean countries, Poland's State Forests (Lasy Państwowe, the state body managing the country's forests) have for some time been testing similar solutions, deploying AI systems to detect fires based on camera and satellite imagery. Galicia's example shows that, with sufficient investment, camera networks powered by convolutional neural networks can genuinely cut emergency response times from hours to just minutes, something that, as heat waves and drought periods grow more frequent in Central Europe, could become standard in Poland too in the coming years.

The experiences of Galicia, Greece, and Portugal could also serve as a reference point for planning EU wildfire protection funding, given that the record 2025 season affected nearly every EU member state, not just Mediterranean regions.