What We Are Building and Why

We presented Pyr-Stop publicly for the first time in a more formal context this autumn. This post is a plain-language version of that presentation — what we are building, why we think it matters, and where we are headed.

The Problem We Are Solving

Wildfires are getting worse. The 2025 European fire season was the most damaging since reliable records began, with well over a million hectares burned across the EU. A significant portion of that was in protected natural areas — habitats that cannot simply be replanted or restored.

This is not a static problem. Climate conditions are making large fires more frequent. The fire-weather window — the period each year when conditions are extreme enough to generate fast-moving, difficult-to-control fires — is lengthening.

The core operational challenge is time. A fire discovered at ignition is manageable. A fire discovered two hours later is often not. Conventional firefighting assets — aircraft, ground crews — have mobilisation times, range constraints, and operational limits (including the inability to fly at night) that mean there is almost always a gap between ignition and effective response.

That gap is where the damage is done.

What Pyr-Stop Does

Pyr-Stop is designed to close that gap with autonomous detection and suppression UAVs, coordinated by a ground control station and operating under continuous human oversight.

The system pipeline is: Detection → Verification → Intervention.

UAVs equipped with visible and infrared sensors continuously monitor their assigned areas. When a thermal or visual signature consistent with fire is detected, the system moves to verification — the AI analyses the event in more detail, and a physics-based fire propagation model assesses how the fire is likely to develop. The human operator reviews the output. If confirmed, suppression UAVs are deployed to the fire front.

The system is designed specifically for night-time operations — closing the gap that currently exists between dusk, when manned aircraft stand down, and dawn. It is also designed for terrain where helicopters and fixed-wing aircraft cannot operate effectively.

The Human-on-the-Loop Model

We are sometimes asked whether Pyr-Stop is "fully autonomous." It is not — and that is a deliberate design choice.

A human operator monitors all active missions from the Ground Control Station. The system surfaces information, flags conditions, and can act rapidly — but the operator has visibility throughout and can intervene. We call this the "human-on-the-loop" model.

We believe this is the right approach for high-stakes autonomous systems operating near people. It is also the approach regulators and partner organisations expect.

XPRIZE Wildfire

Earlier this year, Pyr-Stop reached the Semifinals of the XPRIZE Wildfire competition — a global challenge to accelerate autonomous wildfire response technology. The experience validated our core technical approach and gave us meaningful insight into the real operational requirements of fire agencies and the conditions that matter most in evaluation.

We are no longer competing in XPRIZE. We continue to develop the system independently.

What Comes Next

We are in active conversations with agencies, utilities, and land managers about early pilot deployments. A pilot means working closely with a partner organisation — a fire service, a grid operator, a forestry agency — to deploy the system in a real operational context with agreed safety requirements and evaluation criteria.

If you represent an organisation that manages land or infrastructure in fire-prone areas, we are interested in talking. Get in touch.