Emerging Tech

A.I. cameras could help stomp out wildfires before they become disastrous

This summer marks one year since California’s deadliest and most destructive wildfire season ever. And while not for a second diminishing the devastation caused by that disaster, it’s not an isolated event. Many countries around the world are experiencing unprecedented heat waves, which pose similar fire risks. Could cutting-edge technology help?

The folks behind Bee2FireDetection certainly believe that it can. They’ve developed technology, in conjunction with IBM, that can help spot burgeoning wildfires earlier. This can, in turn, help firefighters better distribute resources and take proactive steps to ensure that small sparks don’t turn into larger blazes.

“Early detection is key,” Vasco Correia, chief business officer of Bee2FireDetection, told Digital Trends. “If you’re fighting a fire in its first 20 minutes, it’s most likely that you won’t end up facing huge, devastating fires. If you wait longer than that, you’re simply trying to minimize damage and contain a situation that is already a disaster.”

The Bee2FireDetection system utilizes three different types of camera. The first of these is an optical HD camera, which uses IBM A.I. to identify smoke columns and light variations. The second, a thermal infrared camera, monitors for temperatures that show a fire has either broken out or is about to. Finally, a spectrometric camera can classify smoke columns by chemical component analysis. “We can identify if a smoke column is coming from wildfire, if it’s a cloud that looks like smoke, or if it’s a dust cloud,” Correia said.

ai camera stop wildfire screen shot 2019 07 19 at 18 37

This triple-pronged approach helps avoid potential false positives. While it’s certainly better to think there’s a fire when there isn’t than to miss an actual fire, false positives are still problematic. That’s because firefighters, alerted to a false blaze, could waste time and resources better spent elsewhere.

The system can also pull in weather data from the IBM-owned Weather Company to forecast how likely a fire is in a certain area and how a fire will spread if it does occur. This information can be extremely useful for making key strategic decisions.

“We’re living in a world of change,” Correia said. “With wildfire frequency having increased by 400% since the 1970s, the existing strategies simply were not working.”

Currently, Bee2FireDetection is up and running in Brazil and Portugal. As of now, it’s not being used in the United States, although the company just recently opened its first U.S. office.

Editors’ Recommendations

Let’s block ads! (Why?)

Emerging Tech – Digital Trends

Leave a Reply

Your email address will not be published. Required fields are marked *