Ingkayefe

A system to predict forest fires that combines devices with multiple sensors that send information to the cloud and a predictive model with AI.

Forest fires are a growing threat throughout the world, especially due to climate change in areas where temperatures are expected to increase and rainfall to decrease, such as in the central-southern area of Chile. Only in 2017, 518 thousand hectares were burned. Approximately half were forest plantations, while the other half corresponds to native forest and others, which means great economic losses for companies and also great environmental damage. There are 3 million people who live in flammable forest-urban interfaces, and the number of inhabitants who have already been affected is close to 55 thousand. Mega fires (fires that exceed 10,000 hectares) release large amounts of greenhouse gases and particulate matter, seriously damaging the respiratory system of volunteers and the affected population. Only between 2010 and 2018, 16 of these claims were registered with 444,000 hectares affected. 

Ingkayefe consists of a forest fire early warning and prediction system. One of its main components is a device for the detection of accidents whose purpose is to warn as quickly as possible that conditions in a certain sector are favorable for a fire to break out or that one is already taking place. Its components are an ESP32 board with temperature, humidity and wind speed sensors. The device sends this data to a server which builds a heat map that reflects the probabilities of a forest fire taking into account the information sent by the devices as well as historical information and information related to the topography of the terrain, among other data. . Unlike other alert or prediction systems, ours combines hardware and software to obtain information, process it and display results that are as close to reality as possible. It also has a social component as it integrates an app that alerts users in areas with high risk of a forest fire occurring and educates on how to prevent it. The objective of our project is to prevent large forest fires that may affect forests, forest plantations (and therefore the economy) and rural and urban populations.

It is expected that our project will benefit rural and urban populations exposed to the dangers of forest fires, but not only them, since one of the main objectives is to protect the native forest (a factor of great importance in stopping climate change) and also the fauna that inhabits it, among which are endemic protected species. Economic progress and jobs are also positively affected as it is a step towards a more responsible and ecological forestry industry.

Currently, our "team" is only made up of its leader who is currently studying engineering at one of the most prestigious universities in Chile, the University of Chile. However, in the future it is expected that more talented people will join to work on the project to form a multidisciplinary team connected with the community.

Summer after summer we observe how forest fires become a major problem, affecting people's lives, the economy and the environment. That is why an idea was born that combines different measures that already existed but also includes technology to make them more efficient.

Our project combines 2 existing solutions to fight forest fires: devices that measure environmental variables (wind speed, temperature and humidity) and predictive models. In addition, our devices send their data to the cloud, unlike those currently used, automating a process and making it more efficient.

During the next year we hope to have a finished product to start offering it to what we hope will be our main customers: forestry companies and local governments.

To achieve that goal we are looking for more people to join our team, so that is also one of our goals.

Our devices are programmed with Arduino while the web application it use to store the data runs on PHP.

The predictive model that we plan to develop will be programmed with Python and will use artificial intelligence and machine learning.

In the following year, if we have our product ready, we could test it in our local community, thereby improving the lives of up to 20,000 people.

We do not have funds to pay for the development of our project nor do we have the necessary contacts to market it. In addition, due to the university, we will not have much free time to work.