We all have to do our part in clearing land
As Portugal braces for another demanding fire season, the country is turning to artificial intelligence not as a novelty, but as a means of honoring the hard-won lessons embedded in decades of wildfire records. Secretary of State Rui Rocha envisions a system where historical patterns guide future decisions — positioning people and equipment before disaster strikes rather than merely responding to it. The initiative reflects a broader human reckoning with fire: that in a warming world, anticipation may matter more than reaction.
- Portugal expects summer 2026 to be one of its most challenging fire seasons yet, with officials speaking in tones of urgency rather than optimism.
- The government is racing to integrate AI into its rural fire combat system before peak season, with coordination meetings already underway between civil protection and digitalization officials.
- A practical crisis is simmering beneath the high-tech ambitions: landowners and municipalities are struggling to clear vegetation by legal deadlines, citing labor shortages and overwhelming workloads.
- On the ground, Portugal is deploying its largest fire-fighting force to date — more than 15,000 personnel, 350+ vehicles, and fire retardant now pre-positioned at four air bases instead of one.
- The strategy is landing as a layered bet: AI-driven foresight paired with expanded physical resources, though officials concede that weather remains the variable no algorithm can fully tame.
Portugal is moving to embed artificial intelligence into its wildfire defense ahead of a summer that officials expect to be severe. Secretary of State for Civil Protection Rui Rocha outlined the vision during a visit to the Guarda district, where he was presenting the special rural fire combat device for the Beiras and Serra da Estrela region.
The core idea is straightforward but powerful: Portugal's National Emergency and Civil Protection Authority holds decades of wildfire records — where fires start, how they spread, what conditions feed them. Applying machine learning to that archive, Rocha argued, could help commanders pre-position resources and make faster decisions when fires break out. Coordination with digitalization secretary Bernardo Correia is already underway, with the ambition of having AI tools integrated into the rural fire combat device before the year is out.
A more immediate tension surrounds vegetation clearance deadlines. Set for May 31 in most of the country, the cutoff has drawn appeals from landowners and local governments alike, who cite labor shortages and the scale of work required following recent storms. Rocha said the government would assess conditions by month's end before deciding whether an extension was warranted.
For the critical July-to-September window, Portugal is fielding its largest fire-fighting force yet: more than 15,000 personnel, over 350 vehicles, and 50 tracked machines. Fire retardant — previously available only from Santarém — will now be pre-staged at four air bases across the country, and two Air Force Black Hawks are being repositioned to strengthen aerial response.
Rocha was candid that weather will ultimately set the terms of the season, but expressed confidence in the expanded and increasingly intelligent system being assembled. He closed with a reminder that technology and logistics alone are not enough — prevention, he said, is a shared responsibility that begins with every cleared hillside and maintained firebreak.
Portugal is preparing to deploy artificial intelligence in its wildfire defense strategy ahead of what officials expect to be a difficult summer. Rui Rocha, the country's secretary of state for civil protection, laid out the plan during a Friday visit to Figueira de Castelo Rodrigo in the Guarda district, where he was presenting the special rural fire combat device for the Beiras and Serra da Estrela intermunicipal region.
Rocha's pitch for AI centers on its capacity to absorb Portugal's vast historical wildfire data and use it to anticipate dangerous scenarios before they unfold. The National Emergency and Civil Protection Authority has accumulated records spanning decades—patterns of where fires ignite, how they spread, what conditions accelerate them. Machine learning applied to that archive could help commanders position resources more effectively, move personnel and equipment to the right places at the right times, and make faster, better-informed decisions when fires break out. "The artificial intelligence applied to such large historical data sets as we have can be very important for anticipating and preparing a set of operational scenarios," Rocha told the news agency Lusa.
The government has already begun coordinating with its digitalization secretary, Bernardo Correia, to make this happen. Rocha said they've met once and plan to meet again the following week. The goal is ambitious: to have AI tools integrated into the rural fire combat device by the end of 2026. "We want to pursue this opportunity very much," Rocha said. "We'll see if we can get the device to use AI still this year, but it's something we want to increase more and more, also in the operational part."
Beyond AI, Rocha addressed a more immediate practical problem: the deadline for landowners to clear vegetation from their properties. Currently set at May 31 for most of Portugal and June 30 for areas hit hard by recent storms, the deadline has become a source of pressure. Both private landowners and local authorities have asked for an extension, citing labor shortages and the sheer volume of work required. Rocha indicated the government would evaluate conditions by month's end to determine whether pushing the deadline further into summer made sense. "Many entities and also private individuals have asked that the period should be longer, especially this year, when there is much lack of labor and much work to do," he acknowledged.
The summer ahead will test these preparations. During the critical months of July, August, and September, Portugal's fire combat device will deploy more than 15,000 personnel, over 350 vehicles, and 50 tracked machines—a larger force than the previous year. The government has also expanded access to fire retardant, the chemical dropped from aircraft to slow flames. Last year, retardant was available only from Santarém; this summer it will be stationed at four air bases: Vila Real, Viseu, Proença-a-Nova, and Cernache. Two Black Hawk helicopters from the Air Force will be relocated from Ovar to Monte Real to strengthen aerial capacity.
Rocha acknowledged the season will be difficult and that weather remains the ultimate constraint on any strategy. But he expressed confidence that the expanded device, now fortified with additional equipment and the promise of AI-driven planning, is ready for what lies ahead. He also emphasized that the burden of prevention falls on everyone—local officials, residents, and civil protection workers alike. "We all have to do our part," he said, "in clearing land, which is very important to guarantee the best conditions for fighting fires."
Citas Notables
Artificial intelligence applied to such large historical data sets can be very important for anticipating and preparing a set of operational scenarios— Rui Rocha, secretary of state for civil protection
We want to pursue this opportunity very much. We'll see if we can get the device to use AI still this year— Rui Rocha
La Conversación del Hearth Otra perspectiva de la historia
Why does Portugal think artificial intelligence is the right tool for wildfire management right now?
Because they're sitting on decades of fire data that no human team could fully process. AI can find patterns in where fires start, how weather and terrain shape them, which resources worked and which didn't. That's the kind of pattern recognition that helps you position firefighters and equipment before a fire even starts.
But doesn't that require the AI to be trained and tested before summer actually arrives?
Exactly. That's why Rocha is pushing so hard to get it operational by year's end. They're not waiting for perfection—they want it working in the field this summer, learning as it goes.
What about the land-clearing deadline? That seems like a separate problem.
It is, but it's connected. If people don't clear dead vegetation from their property, fires spread faster and burn hotter. The government set May 31 as the deadline, but labor is scarce and the work is immense. Rocha is essentially saying: we might give people more time, but only if weather conditions allow it.
So the AI helps with response, but land-clearing is about prevention.
Right. One is about being smarter when fire comes. The other is about making sure fire doesn't spread as easily in the first place. Both matter.
Is 15,000 personnel a lot?
It's more than last year. But Rocha was honest—weather is still the thing that decides everything. You can have all the people and machines and AI in the world, but a hot, dry wind changes the game instantly.