dualCore| Spot That Fire V3.0

Spot That Fire V3.0

Recent wildfires worldwide have demonstrated the importance of rapid wildfire detection, mitigation, and community impact assessment analysis. Your challenge is to develop and/or augment an existing application to detect, predict, and assess the economic impacts from actual or potential wildfires by leveraging high-frequency data from a new generation of geostationary satellites, data from polar-orbiting environmental satellites, and other open-source datasets.

AGNI

Summary

Agni is a web app that incorporates geographic and weather data and a machine learning model to predict possible areas of high risks of forest fires and detect ongoing forest fires, with the ability to predict the fire's intensity for deeper analysis of the impact of the fire upon the social-economic status of one or multiple countries.

How We Addressed This Challenge

Since we have in mind the idea that prevention is better than a cure, we tried to approach the problem by giving authorities and communities a resource that tells them where there is a high probability of a wildfire breaking out, to reduce impact upon the human population and decrease the time taken by the appropriate authorities to respond, reducing the economic impact of the fire. It works by taking data from a meteorological API of humidity, temperature and precipitations in the last hour in a geographic area, and that data is run through our Machine Learning model to accurately predict a possible wildfire's extent, or to detect an ongoing wildfire through NASA's satellite data.

How We Developed This Project

This topic seemed very relevant in today's world, considering the Amazon, Australia and California fires, that we feel could've been prevented and contained much faster had the support tech been there. To develop the project we split up our forces as much as possible to cover more ground, since there are only 2 coders in our team of 3. We used React.JS for the front-end, IBM Watson Machine Learning, Firebase Real-Time Database and the MeteoMatics API alongside the MapBox API. We encountered quite a lot of problems, but we're happy to have learned more from the experience.

How We Used Space Agency Data in This Project

We have requested archived data from NASA Firms regarding wildfires that took place over the past year worldwide. Using this we have trained an IBM Watson Machine Learning Model using AutoAI to make a possible prediction on how big the affected surface (taking into consideration human intervention that has been in the past fires in the area). Now in order to predict where a fire could start, we have used meteorological data and we come up with a formula that gives us a probability of happening ( the formula is not perfect, but enhancements of the formula are possible). 

Project Demo

https://www.slideshare.net/RaulFloarea/spot-that-fire-v3

https://we.tl/t-SqbB9bfyMn (pt video)

Project Code
Data & Resources

NASA Firms (https://firms.modaps.eosdis.nasa.gov/download/)

IBM Watson Machine Learning (https://eu-gb.dataplatform.cloud.ibm.com/projects/ea917012-b997-483b-9cbc-8632df4059f9/overview?context=cpdaas).

Meteomatics Weather API to get a chance of fire in a specified region (https://www.meteomatics.com/en/nasa-space-apps-challenge/) .

Judging
This project was submitted for consideration during the Space Apps Judging process.