In the past few months there has been a massive increase in wildfires in Australia and South america. Due to this, we wanted to respond to this with a remedial solution. We thought of a few ways to tackle this, but a mass surveillance system with a series of interconnected data collectors would be an appropriate/effective solution.
What it does
Our system collects data from portable arduino units, attached with sensor modules and a radio transceiver. This then relays this data up to 1km away! This data is parsed and formatted, then sent to a server that will then store this data, along with the unit's id. This data is then used to populate a map, with its coordinates, with markers. These markers are clickable and the data that they have relayed is displayed. The data is also used to produce a heat map, so depending on the data, 'high risk zones' and 'low risk zones' are generated around these markers.
In addition, it is intended for first responders to be able to react to changes in a fire or potential fire. Therefore, when a unit's data surpasses threshold values, a warning text is sent to users who have signed up for notifications for their location unit. The user will receive updates on their local stations values, as well as be alerted when they are at risk.
How we built it
We built the system using three arduino unos, with radio transceiver modules and sensor modules. We then built a flask server to run the webpage, and the backend. We used MongoDB to store all the data collected from the sensors and the user data. We used the google cloud platform, and used the google API for google maps, where we added custom markers and heat maps according to the data we supplied.
Challenges we ran into
The arduino communication was particularly difficult as created the package then sending it, to then be unpacked on the base station was tricky to get working. We also had problems server side supplying the information to the marker when selected.
Accomplishments that we're proud of
That we managed to overcome these difficulties and produce a product that we are proud of, we also managed to incorporate an excellent UI, with good visuals for viewing the data.
What we learned
We learned how to program with C++, we learned how to use radio communication through arduinos. We also learned how to use the google api to layer markers and heat maps. We also learned how to build an interconnected system, that uses JS, PHP ,JQuery, Ajax requests and MongoDB.
What's next for Earthtones
In the future the sensor units can be expanded to house a wider range of information, which would mean it could be deployed in many different locations to collate environmental data, this can now be deployed in areas with high natural disasters to help local responders react.
As well as that, we did develop a machine learning program that predicts the chance of a fire spreading to a certain area, however due to the dataset that we could find, we were restricted in the training capabilites. This lead to a reduced accuracy.