WEBVTT

00:00:03.562 --> 00:00:07.687
2025 kicked off for Los Angeles residents
with devastating fires

00:00:07.687 --> 00:00:12.437
that burned some 40,000 acres
and killed 29 people.

00:00:12.750 --> 00:00:17.187
After these fires, and as wildfires
become more common around to the world,

00:00:17.375 --> 00:00:20.187
prevention, detection and post-fire
recovery

00:00:20.187 --> 00:00:23.187
technology is back on everyone's radar.

00:00:23.250 --> 00:00:26.250
California is very familiar
with these types of solutions.

00:00:26.250 --> 00:00:29.875
The state's Department of Forestry
and Fire Protection, or CAL FIRE,

00:00:30.125 --> 00:00:34.500
has actively been using early fire
detection technology for over a decade

00:00:34.687 --> 00:00:38.562
and even created Alert California,
a program that uses

00:00:38.562 --> 00:00:42.687
AI enabled cameras
to spot, track and monitor smoke and fire

00:00:43.000 --> 00:00:46.375
that has played an important role
in the state's latest fires.

00:00:46.562 --> 00:00:50.000
In 2019, we had, through legislation,

00:00:50.000 --> 00:00:54.500
some funding that would allow us

00:00:54.687 --> 00:00:59.375
to add some additional ground
mounted camera systems

00:00:59.500 --> 00:01:03.562
to the existing network
that Alert California had.

00:01:03.562 --> 00:01:07.937
UCSD and Alert
California had enough data to then

00:01:09.062 --> 00:01:10.187
build a machine learning

00:01:10.187 --> 00:01:13.187
model for change detection,
and that change detection

00:01:13.250 --> 00:01:16.062
was for smoke and fire,

00:01:16.062 --> 00:01:19.437
which then turned
into artificial intelligence

00:01:20.062 --> 00:01:22.562
within the camera system
within the network.

00:01:22.562 --> 00:01:26.500
The program, based at the University
of California San Diego and developed

00:01:26.500 --> 00:01:30.750
with CAL FIRE, utilities and DigitalPath,
has led to the deployment

00:01:30.750 --> 00:01:34.875
of more than 1100 cameras
across the Golden State.

00:01:35.125 --> 00:01:40.000
The devices can allegedly spot
smoke and fire up to 60 miles away

00:01:40.000 --> 00:01:43.375
during the day, and 120 at night.

00:01:43.625 --> 00:01:47.687
When a camera detects
what the AI identifies as smoke or fire,

00:01:47.937 --> 00:01:51.187
it sends an alert and footage
to a control center

00:01:51.187 --> 00:01:56.250
that verifies, validates, and notifies
the emergency command centers of CAL FIRE,

00:01:56.500 --> 00:01:59.500
allowing those in charge to address
the issue.

00:01:59.625 --> 00:02:02.687
Alerts are sent via text message, email,
and a phone call.

00:02:02.937 --> 00:02:06.687
The program has had a funding of over
$20 million.

00:02:07.000 --> 00:02:11.437
Our mission is to suppress
95% of our fires at ten acres or less.

00:02:11.875 --> 00:02:15.437
So what we have seen,
the success rate with these cameras

00:02:15.437 --> 00:02:20.062
and what it has done is
it's really helped us with that mission

00:02:20.062 --> 00:02:23.750
of more than 95% of our fires
at ten acres or less.

00:02:23.750 --> 00:02:27.062
So the value
and the success that we have seen

00:02:27.312 --> 00:02:30.312
are the fires that the AI detection

00:02:30.437 --> 00:02:34.062
has notified us, sent that notification,

00:02:34.312 --> 00:02:38.687
the command center receives
it, dispatches resources to respond.

00:02:38.687 --> 00:02:43.125
They suppress the fire
and then go back without even a 911

00:02:43.125 --> 00:02:46.875
phone call, an emergency phone call
going into those command centers.

00:02:46.937 --> 00:02:50.250
The cameras generally use cellular
or Wi-Fi connectivity,

00:02:50.250 --> 00:02:53.250
but can also rely on satellites in off
grid areas.

00:02:53.500 --> 00:02:57.250
They can be powered by solar, batteries
or be mounted on utility

00:02:57.250 --> 00:03:01.312
infrastructure, as these companies
are partners and sponsor devices.

00:03:01.562 --> 00:03:05.687
But the cameras and Alert California
aren't the only technology

00:03:05.687 --> 00:03:09.375
and programs
CAL FIRE is relying on to put out fires.

00:03:09.625 --> 00:03:13.625
Technology plays a role before, during
and after fires,

00:03:13.625 --> 00:03:16.437
something
we saw during the recent LA fires.

00:03:16.437 --> 00:03:19.312
Some of the pre is our
predictive services.

00:03:19.312 --> 00:03:22.500
And a lot of our predictive services

00:03:22.500 --> 00:03:27.625
products are generated
from a company called Technosylva.

00:03:27.625 --> 00:03:31.312
So they have a wildfire analyst
that we have an enterprise license through

00:03:31.625 --> 00:03:33.562
and that gives us fire risk

00:03:34.500 --> 00:03:36.000
several days out

00:03:36.000 --> 00:03:41.250
that we can identify potential heightened
hazard within specific areas

00:03:41.250 --> 00:03:45.062
throughout the state,
as well as our fire simulation.

00:03:45.437 --> 00:03:47.500
So our fire spread simulation.

00:03:47.500 --> 00:03:51.250
So before the fire happens,
we can see that risk.

00:03:51.250 --> 00:03:54.562
Once the fire does occur
and in combination

00:03:54.562 --> 00:03:59.000
with some of our National Guard assets
for early detection, ongoing assessment

00:03:59.250 --> 00:04:02.250
with the camera detections,

00:04:02.500 --> 00:04:07.125
we coalesce all of those things
to make fire spread predictions.

00:04:07.125 --> 00:04:12.875
And those will then show us,
in a simulated fashion, up to 100 hours

00:04:12.937 --> 00:04:17.687
of what the possible spread or potential
spread of that fire would be.

00:04:17.687 --> 00:04:20.312
So it gives us a good basis
and understanding.

00:04:20.312 --> 00:04:24.500
If the fire starts here,
this is potentially what it could impact.

00:04:24.625 --> 00:04:28.437
The GIS, AI enabled software
prediction system by Technosylva

00:04:28.437 --> 00:04:32.125
allows the firefighters to identify risk
areas.

00:04:32.375 --> 00:04:36.187
The platform runs simulations
thanks to camera, weather, vegetation

00:04:36.187 --> 00:04:37.375
and soil moisture,

00:04:37.375 --> 00:04:41.875
topographical, historical, infrastructure,
and Earth observation data.

00:04:42.250 --> 00:04:45.875
After running the simulations,
firefighters can see if the area

00:04:45.875 --> 00:04:49.687
they're working in is a high spread zone,
the infrastructures or properties

00:04:49.687 --> 00:04:53.500
that will be affected to decide
where priorities will lie.

00:04:53.812 --> 00:04:57.375
After all the data is analyzed
on the Technosylva software

00:04:57.375 --> 00:05:01.187
in their common operational platform,
they can also see information

00:05:01.187 --> 00:05:05.875
on jurisdictions, population, waterways,
utility infrastructure...

00:05:06.312 --> 00:05:11.062
All the technologies used
are pieces of a puzzle that, put together,

00:05:11.125 --> 00:05:14.812
allow for a better understanding
of the situation on the ground.

00:05:15.312 --> 00:05:18.750
I think that what we have done, interwoven

00:05:18.750 --> 00:05:21.375
all of these pieces of technology,
whether it's fire

00:05:21.375 --> 00:05:24.812
guard for the early detection,
ongoing assessment or incident

00:05:24.812 --> 00:05:29.437
awareness and assessment,
our fixed wing platforms, our aircraft

00:05:29.437 --> 00:05:35.312
that are pulling that infrared perimeters,
in near to real time.

00:05:35.312 --> 00:05:40.875
The cameras looking at where
that column of smoke is laying over,

00:05:40.875 --> 00:05:44.312
what is it
impacting, the fire spread simulations.

00:05:44.312 --> 00:05:50.812
Adding all of those things into our Fire
Management Assistance Grant, so our FMAG,

00:05:50.812 --> 00:05:54.500
which then opens the door
for additional funding

00:05:54.562 --> 00:05:57.562
to ensure
that those communities are resilient.

00:05:57.875 --> 00:06:00.687
Drones were also used during the LA fires,

00:06:00.687 --> 00:06:04.000
but sometimes struggled due to the heavy
Santa Ana winds.

00:06:04.250 --> 00:06:05.125
After the fire,

00:06:05.125 --> 00:06:08.687
there are still threats that CAL FIRE
are assessing with technology.

00:06:08.875 --> 00:06:12.875
The post-fire threats are mudflows,
debris flows, floodings...

00:06:13.125 --> 00:06:17.562
CAL FIRE is relying on catastrophe
risk modeling software Katrisk for flood

00:06:17.562 --> 00:06:22.125
monitoring to ensure those affected
are safe from the aftermath.

00:06:22.437 --> 00:06:28.000
California has 163,696mi²,

00:06:28.000 --> 00:06:31.875
or 423,970km².

00:06:32.250 --> 00:06:37.125
It's impossible to cover every inch of
the state with cameras to monitor fires.

00:06:37.250 --> 00:06:41.312
The system allows CAL FIRE to contain
fires and keep them from spreading,

00:06:41.312 --> 00:06:45.625
but when early detection fails
due to absence of cameras in an area

00:06:45.812 --> 00:06:49.437
or due to a rapid spread
because of drought or strong winds,

00:06:49.687 --> 00:06:51.937
it's key
for them to have situational awareness,

00:06:51.937 --> 00:06:56.062
which is made possible thanks to
AI driven software models they run.

00:06:56.375 --> 00:07:00.312
And the LA fires
will only reinforce this concept of using

00:07:00.312 --> 00:07:04.250
AI driven solutions to help fight
the great threat wildfires pose,

00:07:04.500 --> 00:07:09.312
especially in the US'
most populous and most fire prone state.