Artificial Intelligence a way to Predict Natural Disasters
Recently, Sonoma County in California has decided to enter a contract with a South Korean technology firm, Alchera. This partnership will give the county an extra set of eyes to help alert for danger of wildfires embodied by drones. When they spot a fire or signs of one, the drones will alert first responders.
The two-year contract was funded through 3 Millon dollars destined for hazard mitigation grants that the Federal Emergency Management Agency awarded to Sonoma County. The funding is also to be spent on 27 extra cameras that will be added to the state’s AlertWildfire system according to Emergency Management Director Chris Godley.
In the midst of installations and configurations Godley said that “emergency responders will have to spend time teaching the AI system how to discern between smoke from geysers and smoke from a fire. This adaptation period may last until Fall, when the system is schedule to be fully operational”.
According to different sources, Alchera’s systems learn by sending an alert with a picture to a dispatcher when it detects smoke. The dispatcher enters a simple “yes” or “no” and thus teaches the artificial intelligence to understand whether that picture is from a real fire or not.
These systems are expected to reduce between 20 and 30 minutes and sometimes even up to an hour in response time, which in a wildfire makes a world of difference.
According to the National Interagency Fire Center: “2021’s wildfire season is predicted to be another severe one. From January 1st to June 14th there have been 33,395 fires in the United States, 9 thousand more than in 2020 already”.
Along with Alchera, several other similar applications have appeared. The IQ Fire Watch in Germany is one example. This application is trained in the Brandeburg National Forest, to detect smoke.
Another is Insight robotics which, based in Hong Kong, installed networks of thermal cameras to detect Wildfires in Spain, Portugal, Mexico and China.
Finally, Omdena is a global technology firm based in Palo Alto, that partnered with a Brazilian Agriculture tech provider, Sintecsys. The developed an artificial intelligence that could identify images of real fires with an accuracy of about 80%.
According to a report by Swiss Re, wildfires have caused a cumulative $56.3 billion in insurance losses from 2011 to 2020. By comparison, wildfires caused a cumulative $8.7 billion in losses from 2001 to 2010, $5 billion from 1991 to 2000 and $1.4 billion from 1981 to 1990. That in an increase of 7.3 billion in 20 years!
The uses of artificial intelligence in insurance vary from risk prediction to risk assessment, but in this case Artificial Intelligence can actually help mitigate the damages from a natural catastrophe.