The Problem is Wildfires Destroy Livelihoods
There exist many natural disasters in the world. Flooding, fires and tornadoes all cause huge widespread damage and suffering worldwide, affecting people and infrastructure. Today, in the 21st Century, we can predict rainfall and thus calculate the chances of flooding, we can monitor wind speeds and directions and thus predict hurricane paths. And yet, there seems to be very few resources for predicting wildfires, particularly with regard to using up and coming technologies like artificial intelligence and machine learning.
Our Solution is Predicting Wildfires Using Artificial Intelligence
We predict where outbreaks of wildfires are most likely to occur, based on hyperlocal weather data. Our prediction model is based on advanced machine learning techniques, giving an accuracy of 74.8%. Using our predictions, firefighters worldwide can look ahead and prevent wildfires before they happen.
Our team is made up of Computer Science and Maths students from the University of York. Peter and Vishal are both in the placement year of their four year Computer Science degrees. Flinn graduated from the University of York with a degree in Mathematics. He has recently embarked on a Masters at University College London.
University of York students Peter, Vishal and Flinn joined forces and began development of Wildfire AWARE, aiming to compete in the NASA Europa Challenge.
The team spent several days in Helsinki, Finland, competing in the NASA Europa Challenge Finals. Judged by a team of NASA, ESA and Finnish researchers and experts, Wildfire AWARE came in 4th place.
AWARE was awarded First Prize in the SMARTIES Competition, a competition designed to promote innovate smart city applications, funded by the Research Councils UK.
Wildfire AWARE is still in development. We are constantly finding new ways to increase the accuracy of our predictions and improve the way we visualise this data on the web. The future is bright for AWARE!