Previously, we had already decided to take part in the Spot That Fire V3.0 challenge.

Initially, we created a working Board on miro.com. In it, we performed a first brainstorming to support us in defining which problem would be addressed in our project. In our research, we found that it was imperative to offer an information tool that could speed up the response time of public agents in combating wildfires. In addition, it was necessary to provide the population with alerts as updated as possible, with recommendations for procedures in the event of fires with the potential for rapid spread.

Then, we defined the personas that would be served -- public agents and citizens/companies, as asset registrants --, as well as the types of alerts that would be the issued by our online tool: a) wildfire season started, b) new fire outbreak, c) fire with low risk of spread, d) fire with high risk of spread, e) critical proximity alert, and f) periodic reports.

So, we started the development using these technologies: a) Supervised machine learning with SVM - Support Vector Machine, b) Adobe XD, c) Java, d) Python, e) Dockerfile, f) HTML, g) Javascript, h) CSS/SCSS, i) Typescript, and j) Angular.

We used NASA FIRMS Active Fire Data* to train the machine learning / SVM model, and to recover near real time data.

In our MVP, we also use climate data from INMET*, the Brazilian National Meteorological Institute, to complement machine learning.

* See Data & Resources section below.

Presentation:

https://drive.google.com/file/d/1YU2RkaUyyMw3oelSSoBplaD3EAUHRkVJ/view?usp=sharing

Smokejumpers - Skope - Pitch 30 seconds english final

https://youtu.be/_NqXFC5BA6s

Smokejumpers - Skope - Pitch 3 minutes portuguese final

https://youtu.be/cOfbe9Ur3zM

Data sources used by machine learning:

https://firms.modaps.eosdis.nasa.gov/download/

https://portal.inmet.gov.br/dadoshistoricos

Data sources used by application:

https://community-open-weather-map.p.rapidapi.com/weather

https://openweathermap.org/current

https://nrt3.modaps.eosdis.nasa.gov

References:

1.   A Data Mining Approach to Predict Forest Fires using Meteorological Data <http://www3.dsi.uminho.pt/pcortez/fires.pdf>

2.   Programa Queimadas - Instituto Nacional de Pesquisas Espaciais <http://queimadas.dgi.inpe.br/queimadas/portal>

3.   Quase 500 bebês foram internados por conta de fumaça da Amazônia em 2019 <https://noticias.uol.com.br/meio-ambiente/ultimas-noticias/redacao/2020/08/26/amazonia-queimadas-internacoes-criancas.htm>

4.   NASA - Earth Observatory - Fire <https://earthobservatory.nasa.gov/global-maps/MOD14A1_M_FIRE>

5.   Influência dos Elementos Meteorológicos Sobre o Comportamento do Fogo <https://www.scielo.br/scielo.php?script=sci_arttext&pid=S0102-77862019000100033&lang=pt>

6.   World Health Organization - Wildfires <https://www.who.int/health-topics/wildfires#tab=tab_1>

7.   Forest Fire Prediction <https://www.knime.com/knime-applications/forest-fire-prediction>

8.   Incêndios Florestais: o risco sempre iminente <https://www.matanativa.com.br/blog/incendios-florestais-o-risco-sempre-iminente/>;

9.   Incêndios Florestais: Causas, Consequências e Como Evitar <http://www.ibram.df.gov.br/wp-content/uploads/2018/02/Cartilha-Inc%C3%AAndios-Florestais-Causas-Consequ%C3%AAncias-e-Como-Evitar.pdf>

10. ANEEL inicia nova campanha de prevenção a incêndios <https://www.aneel.gov.br/sala-de-imprensa-exibicao/-/asset_publisher/XGPXSqdMFHrE/content/aneel-inicia-nova-campanha-de-prevencao-a-incendios/656877?inheritRedirect=false>;

11. ANEEL - Desligamentos de linhas de transmissão provocados por queimadas < https://www.aneel.gov.br/documents/656835/14876457/2018_DesligamentosLinhasTransmissaoQueimadas.pdf>

12. Verisk - Wildfire Risk Insight Analysis of property exposure and wildfire damage in 2019 <https://www.verisk.com/siteassets/media/campaigns/gated/underwriting/2019-wildfire-analysis.pdf>