Problem:

One billion, maybe when we hear this number we think of a unicorn or a huge well-established company, unfortunately, this number refers to how many animals vanished due to the wildfires in Australia last year. 32 species of them are in danger of extinction.

Wildfires are one of the worst disasters to hit country. It destroys everything not only trees, but also humans, undermines national economy by billions of dollars and destroys the bio-systems.

All of these effects affect sustainable development and prevent us from achieving this goal as it destroys the 3 main pillars of sustainable development which are; Environmental pillar, Economical pillar and Social pillar.

An average of 72,400 wildfires cleared an average of 7 million acresof U.S. land each year since 2000, double the number of acresscorched by wildfires in the 1990s. In 2015, the largest wildfire season recorded in U.S. history burned more than 10 million acres of land.

Wildfires devastate anything in their path. In 2018, the most destructive California wildfire of all time caused 85 deaths and was the world’s costliest single natural disaster that year with losses exceeding $16 billion.

Causes:

Natural

Only 10 to 15 percent of wildfires occur on their own in nature due to:

·      Vegetation, type of trees and fibers can affect the rate of wildfires as there are types of fibers highly flammable and combustible.

·     Wind, wind speed and direction help us know the rate at which fires spread and help predict the fires behavior. California wildfires are often made worse by the hot, dry Santa Ana winds, which can carry a spark for miles.

·      Geology, earth’s physical structure and substance affect the rate at which fires spread.

·      Temperatures, proportional relationship so when temperature arises, the likelihood of wildfires increases.

·      Humidity, inverse relationship so when humidity decreases, the likelihood of wildfires increases.

Human

·      Population density, electricity networks, houses distribution and carelessness with fire

Vision:

We aim to help the world predict wildfires before happening accurately and save every single human from death in a wildfire.

Our Solution:

Before wildfires

After our climate change and sustainable development student team member, Hajer, had explained wildfires causes and given us a solid understand of them, we began exploring data available from NASA and JAXA, and we found that we could not only hack a lot of wildfires causes using them, but also detect and predict wildfires.

When a change happens in the climate (temperature, wind speed and humidity) of any region on Earth, NASA and JAXA satellites record it, and using some mathematical and agricultural equations we can integrate these data with the nature of trees and geology in this area to determine how likely a wildfire would happen there and how fast it would spread. But how will we do this?

First of all, Yassmin, our data scientist team member, collected data from FIRMS to process them and train a Machine Learning model to classify and detect fires. Then, using data from FIRMS, Landsat 8 and JAXA websites we observe changes in factors lead to wildfires and integrate them automatically with remote sensing and Geographic Information System (GIS) to generate Normalized Differences of Vegetation Index (NDVI) to send these data to our ML model to determine the likelihood of a fire to happen at this area.

According to the likelihood we set the risk on a 3 degree-index, 1 refers to green, 2 refers to yellow and 3 refers to red. For green we don’t send notification as it refers to safe mode, for 2, Karim, our Python developer, programmed automation system to send e-mails to near users, and for 3 our system sends SMS to near users and notification to Authorities.

During wildfires

During the wildfire, our users can use our mobile app (Hot Spot), developed by Esraa our Android developer, to capture images to inform how big the wildfire becomes, and a built-in image processing feature will immediately check the image and verify it. Then, image would be uploaded to our data base to be processed and classified to get more information about the wildfire in this area and send proper notification to near users and authorities and display this info on the map in our app.

Users can release an evocation to get help from other users or authorities, and this evocation would contain the location of this user and a message to explain the situation. This evocation would appear on the map and whoever can help can chat with user in need to give a hand, also they can send a message to hospitals or fire fighters to find them.

To encourage more people help each other, a pints system would be integrated to give points to those who help and then using Corporate Social Responsibility (CSR) to reward them.

Augmented Reality (AR) would be used to guide users to the nearest shelter which would be verified and entered to the app by authorities to be reliable. 

After wildfires

Almost all injuries during fires are skin burns whose syndromes can be captures using image processing as all of them are on the outer layer of skin. So, we use image processing and AR to help injured people know from which degree their skin burn is and how to decrease their pain until paramedics appear.

Wildfires also spawn their own weather systems, including pyrocumulonimbus clouds which NASA calls the “fire-breathing dragon of clouds” for the thunderbolts they hurl at Earth, fueling further blazes and sometimes even fire tornadoes.

Smoke is full of carbon dioxide, carbon monoxide, and particulate matter. It can also contain other chemicals depending on what’s burning. During a wildfire, about 90 percent of the particles in the smoke are smaller than roughly a thirtieth the diameter of a strand of hair, these particles, identified as PM2.5 are so fine, they’re capable of entering lungs and even bloodstream. When looking at an air quality map to determine whether it’s healthy to go outside during wildfire season, you’re largely looking at the concentration of these microscopic particles in the air.

Using API we integrate the Air Quality Index (AQI) map, developed by the Environmental Protection Agency. Which has a range between 0 and 500, 50 is good, which means the air quality is satisfactory, and between 50 and 100 is acceptable air quality. Above 100, the air is classified as “unhealthy for sensitive groups”, and above 150 threatens some members of the general public. Above 200 presents a risk of health effects for everyone. Air quality during the West Coast wildfires was recorded above 500 in some areas, which is simply deemed as “beyond index.” so users can get information about the quality of air before being exposed to it.

Mobile App Overview: