Predicting Fires

Fion’s proprietary algorithm and data sources work together to do what’s never been done—predict where fires will start and how they’ll spread.

From data to prevention

On the cutting-edge

We’re building the first machine learning model with accurate fire prediction & detection, spread prediction and destruction estimation.

We use 12 data sources with real-time machine learning to offer the most accurate fire prediction solution on the market.

Other solutions take days to give you accurate information—Fion does it in hours.

our data sources
Satellite imagery
Sentinel 2 L1C data
Using Sentinel 2 L1C data, Fion’s U-Net architecture assesses satellite images to quickly detect fires and derive more data.
Vegetation indices
Sentinel 2 L1C data
Fion derives vegetation indices like NDMI and NDVI from satellite imagery to understand fuel sources. Using this information, Fion recommends areas for controlled burns and can predict where fires will start and how they’ll spread.
Using data from the National Oceanic and Atmospheric Administration, Fion accounts for wind, pressure, humidity, cloud coverage and temperature.
SRTM data
Utilizing a digital elevation model, Fion assesses topographic gradients for more precise predictions.
a new approach


Current fire detection methods are innacurate and costly, relying on humans or hardware to notify local authorities.

Fion’s U-Net architecture allows it to assess satellite imagery and other data sources to detect fires quickly and accurately.

We’re building a future where fires are stopped before they start.

From data to prevention

The Fion Dashboard

Forget data scientists—Fion is designed to simplify complex data into actionable information and guidance.

Identify at-risk areas
Fion notifies you of locations that are most at-risk for wildfires and tells you how to respond.
Real-time data monitoring
Fion assess data in real-time giving you the most accurate fire prediction and detection insights.

Learn more

How does Fion differ from other solutions for wildfires?

What is Fion developing?

Preventing Fires

With double the data sources and an improved machine learning architecture, we built a new solution for wildfires. One that can predict where fires will start, detect when it starts, predict where it will spread and estimate the damage it will cause.