Project in Spanish

https://drive.google.com/file/d/1D6niifbORpbIRSWMkJVo2KewXnFGiZz7/view?usp=sharing

Forest fire is fire that spreads in an uncontrolled manner in a natural area. Currently, this type of catastrophe affects both the biodiversity of the environment and the surrounding populated areas, and it also generates changes in the environment that can lead to other catastrophes in the future.

Due to their magnitude, their great volatility and the terrain where these accidents occur, it is very difficult for firefighters to detect them, arrive on time and control them quickly. These fires put personnel at great risk; and when making decisions, a mistake can generate many losses, both in material and human terms. As an example, we can mention the 146,000 hectares affected in the province of Córdoba, Argentina.

It is for all this that the challenge of “developing and / or increasing an existing application to detect, predict and evaluate the economic impacts of real or potential forest fires through the use of high frequency data from a new generation of satellites seemed important geostationary, polar orbiting environmental satellite data and other open source data sets "

Our challenge is focused on an application prototype (which uses an API that we have developed) that will serve as a friendly and accessible source of information for brigades to devise a faster and more effective action plan.

Datasets with satellite information are available on the internet, but they are not always easily found. That is why we warn the need to think of an application prototype that automatically acquires, processes and recommends the data and thus ensures that firefighters and brigades can prematurely resolve their actions in order to mitigate significant losses.

 The application uses a genetic algorithm that processes this information, evaluates and suggests the order in which the fire sources should be attacked in order to minimize time, costs and damage. We also have a supervised learning algorithm for classification and an unsupervised learning algorithm for regression of fire grades.

Our challenge focuses on developing an application prototype that serves as a source of information accessible to brigade members so that they can come up with an action plan.

Many of the informational resources are available on the Internet, but they are not always easily found. That is why we are warning of the need to think of an application prototype that automatically processes and merges the data and thus ensures that firefighters or brigade members can resolve their actions in advance in order to mitigate future losses.

The app (built from our API) takes satellite data, processes it and returns it in a friendly interface for the user (brigadistas and institutions) so that they can monitor the variables from a tablet, a computer or a cell phone. The app uses a genetic algorithm that processes and evaluates this information, and suggests the order in which the fire sources should be put out, so as to minimize time, costs and damages. We also have a supervised learning algorithm for classification and an algorithm for unsupervised learning for regression of fire grades.

Centralized information, alerts, processed data that are stored in an application that monitors in real time the variables that brigade members face and that can also provide possible courses of action to be able to act more efficiently and safely contribute to better responses in the fight against forest fires, thus the protection of human life, populations and biodiversity.

Among the feasible solutions generated by our resource and as already mentioned, there is the possibility of using the information to determine areas prone to fire and address the problem early through the use of drones and sensors on the ground. These devices would patrol inhospitable areas and collect data from wide geographic areas. They would serve as radio links. It is worth clarifying in an emergency instance that the data supplied by the satellites has a delay of approximately 1 day that is why it is essential to have application resources that minimize time.

Our proposal became relevant since we had the possibility of contacting the Chief of the Fire Department from Laguna Larga, Cordoba, who gave us interesting information on the subject. Here we summarize the most important:

• What are the main problems you have when going to put out a fire?

 It depends on the type of fire, but if we talk about forest fires most of the time it is the access to the places where the fires are found, mainly due to the topography of the land and on the other hand it is the call of the firefighters for the emergency on weekdays, when they are small claims with few personnel, they are controlled, but for large situations, all personnel must be summoned and on weekdays, due to the personal work of each one, the massive summons of personnel is somewhat complicated.

•  What information do you think it would be good to have on hand before going to put out the fire?

The pre-operational protocols are very useful in an emergency and the constant survey of the coverage areas.

• How is the action plan against a forest fire? Do they prioritize zones?

 Yes, in an action plan lives are prioritized over goods, but it is always around priority sectors, in addition to the action plan there are many more points such as logistics, water recharge, personnel replacement, rest, threatened areas, combustible material, weather conditions at the time. Already in the future, but one of the main points is to be able to read the fire in time and form to be able to evaluate the behavior of the fire.

• What information could you use when they are putting out fires but which is still not information they have in real time?

 Satellite photos updated at the time of the fire since, as they are not continuously updated, the fire load of the place or the accesses, water courses, also fast meteorological information and the place, since a general data for example SMN rarely works, since at the site the properties of the fire mean that these conditions have to be taken locally in order to work more effectively and safely.

In Argentina, firefighters work without being paid for it, that is, they do not receive a salary or state subsidy, they must have a parallel job to support their family, even if they put their permanent life at risk to save other people.

Unfortunately, their resources are scarce, the times they need to buy new equipment they turn to community charities or seek donations.

Our project seeks not only to help the brigade members, but also the population. The information and technology we use is freely accessible to all, allowing people from all over the world to use it and contribute.

We know that this project will help save wildlife, protect the ecosystem and take care of humanity as a whole