A system to empower the wildfire crew by automating the weather observation process
Thanks for the support of Paul Albee, Jenna James, Jay Thiagarajan from Allen Institute for AI.
Thanks for the support of Angie from the Washington State Department of Natural Resources.
And thanks Natallia for designing the nice logo.
What's the Challenge?
Fire forecast is critical for saving lives, but it is difficult
“The forecast is for the fire activity to moderate as well, so it would be unexpected to see that much fire growth,” said Chris Waters, Fire Behavior Analyst .
Weather is the most variable element when anticipating fire behavior
Up-to-date weather information can be critical to fire-fighting agencies.
Accurate forecasts of wind direction and speed help incident commanders make the best decisions to contain wildfires.
Manually weather observation near fireline is time-consuming and risky
* Fire weather is the use of meteorological parameters such as relative humidity, wind speed and direction, mixing heights, and soil moisture to determine whether conditions are favorable for fire growth and smoke dispersion.
What solution did we build?
A systematic solution for automating the collection of real-time weather data near wildland fires
How we arrived the approach?
As wild is a new area for us, we need to learn a lot about the problem scope. We followed the Discover-Define-Prototype-Evaluation process and did rapid prototyping and evaluation of our design for 2 rounds.
Initial Design Question
How might we help wildland firefighting command crew to build robust and low-cost devices for automatic on-site weather data collection?
Ask the right questions
Who are our users?
Who is going to interact with our device?
Who is going to use weather data?
What’s their current workflow?
How do they collect weather data now? What instruments are they using?
How do they need to use the data?
What’s wrong with the current solution?
What are the pain points in their workflow?
How, when, and where will they deploy our device?