Quality wildfire data today will drive the technology of the future
"There's a huge frontier of artificial intelligence, the benefits that it can bring, and it all starts off with good quality data. That's an opportunity we all have, is to start building that database today."
Overwatch Imaging Gregory Denton has to say about the importance of generating quality wildfire intelligence today to drive firefighting technologies of the future.
"I think data is definitely our friend. You know, we look at services like predictive modeling and fire behavior analysis and it’s a very powerful concept and we see the benefits of that in daily operations."
"Where is the fire going to be and how vigorously will it burn? You know, that informs our resourcing, plans and strategic decisions and the aerial resources that we bring to that fight. All of that is predicated on a generation of observations which have taken the collective industry 30 or 40 years to develop. All those models are developed from empirical observations."
"Right now, we kind of lack the next generation observations that are essential for good artificial intelligence and good predictive services. To take the next leap into the future, we all need to start thinking about high quality data observations today. You know, folks at academic institutions are clamoring for it."
"Everyone understands that there are no good observations for fire power and fire energy by rate of output or rate of spread. Smoke columns convective energy being released and the kind of coupled model interactions that go along with fire spread energy release. So, there’s a huge frontier of artificial intelligence, the benefits that it can bring, and it all starts off with good quality data."
"So that's an opportunity that we all have, is to start building that database today."
"When we're talking about data, I think it's more appropriate to say intelligence or information. It’s really what we’re assessing here. Pixels and points and thermal raster’s coming from cameras get translated into a single lat/lon. You know you have a fire confirmed, located here. And that's what we need to do as providers."
"And you can do that through edge processing. So, pushing processing to the aircraft, out into the field so that we're not generating terabytes of 1’s and 0’s that need to be stored in the cloud. Overwatch does a good job of that as well. Rather than this manual process of image interpretation, you can leverage technology to say, your hotspots are here. The active regions of the fire are here and reduce that operator workload, but also the need to store data. If there is a big science data project, you can ship that off to the universities and they’ll have a field day with it. But for the operational community, we need to be better about synthesizing that into actionable intel."