SAN RAMON, Calif. (KGO) — Artificial intelligence is quickly becoming part of our daily lives. We talk to smart speakers to get weather updates, chat bots answer our customer service questions, and soon, self-driving taxis will begin operating in San Francisco. Scientists are also using supercomputing and artificial intelligence to figure out the best way to fight the growing threat of wildfires in California.
“That fast response is very important,” said Ilkay Altintas, the founder and director of WIFIRE Lab at the University of California, San Diego. “When an ignition happens, where the fire is and how to respond to it is vital information for decision support.”
The WIFIRE team uses supercomputers to predict the possible behavior of a fire by analyzing data collected by remote cameras, aircraft, satellites, weather stations and LIDAR sensors.
Firefighters can access the data through a tool called Fire Map. It provides easy access to information about wind speed and direction, humidity, and even the type of vegetation located in the area to determine the fuel potential of a fire.
“This shows we can bring the data together on the initial phases of a fire, locate the fire as accurately as possible and also use the information to predict where the fire will be in the next couple of hours so many decisions can be made,” said Altintas.
Fire Map has been used by firefighters in over 100 fires so far.
These type of tools will prove vital as climate change makes wildfires worse.
A study by researchers at the University of California, Los Angeles predicts Southern California will see an increase in megafires with 25% more land burned by 2040.
“One of the big aspects of artificial intelligence is predictive analytics. It’s using big data. It’s using machines to make calculations and using algorithms to predict certain outcomes,” said Tom Mescall, who heads the Growth and Innovation Team at the consulting firm Armanino.
Mescall works with companies that are interested in deploying artificial intelligence to make their companies more efficient and predict customer satisfaction.
He estimates that by 2025, 95% of all customer interactions with businesses will be with through a virtual assistant powered by artificial intelligence.
The deployment of 5G technology is expected to accelerate the pace of the data revolution as monitoring systems at the core of data heavy technologies, like self-driving vehicles, are able to process information at much faster speeds.
“Monitoring is a big aspect of the Fourth Industrial Revolution. Artificial intelligence brings the ability for computers to learn from the data collected and not only process algorithms, but continually learn over time and get better,” said Mescall.
5G technology is also improving the real-time monitoring of forests and wildlands.
It is also allowing the WIFIRE team at the San Diego Supercomputer Center to develop the next-generation of fire mitigation models.
The team is already testing BurnPro 3D, a platform that uses artificial intelligence to analyze data and help fire managers plan for prescribed burns, a key tactic to reduce the fuel load and the potential of massive wildfires.
“We are building fire models to understand if you had a fire here, what does it consume?” said Altintas.
The team created the first 3D data sets that improve the estimation of fuel and vegetation buildup in the environment.
Up to now, this type of modeling analyzed data at 30 meters resolution in two-dimensional technology.
BurnPro 3D is zooming in to one-meter resolution, a 900% improvement in detail.
Altintas expects the technology to improve to the point where modeling can account for individual pine needles falling on the forest floor.
“We can manage the environment in a way that when a fire happens, it’s going to be manageable,” said Altintas. “That’s the societal promise for positive AI.