-
Tree-mendous, save the trees
Inspiration
Our project aims to end illegal deforestation. In the 26 years between 1990 and 2016, the world lost 1.3 million square kilometers of forest. We have lost 17% of the Amazon rainforest. We are cutting down the lungs of the Earth, limb by limb. While it is difficult to combat deforestation when it is done legally, to clear farmland, or in service of large capitalist enterprises, we can try to stop illegal deforestation. We can surveil the forests and catch deforestation faster. This should make the capture and punishment of the perpetrators easier, which in the long run, should disincentivize the felling of these precious natural resources.
What it does
It detects deforestation in a forest and sends an alert to the authorities using a GPS system
How I built it
The project uses an Arduino Uno and a Raspberry Pi. They will both be housed in a box, attached to a tree. The Arduino will be connected to a microphone, and will use its built-in LED, and the Raspberry Pi will be connected to a GPS module, a Wifi module, and a light sensor on a breadboard. When the microphone on the Arduino senses a chainsaw, it will activate the LED, which will trigger the light sensor on the Raspberry Pi. When the light sensor is triggered, it will use the Wifi Module to send an email to the Forest Service, which will include the coordinates picked up by the GPS. We believe this is an appropriate and effective use of technology, given that all the components are relatively inexpensive, small, and easy to procure, meaning it will not be difficult to widely implement it.
Challenges I ran into
We tried to use machine learning to analyse the frequencies and train the machine to recognize a chainsaw noise. However, we found this to be more difficult than our skill set allowed, so we shifted gears and manually calibrated the range of the chainsaw decibels.
Accomplishments that I'm proud of
Our projects introduces several new approaches and perspectives. Our GPS and Wifi work to send emails, to any email address. Instead of merely setting a decibel threshold, our code for our microphone, analyzes for how long the abnormal decibel levels are happening, and our threshold is a range set by listening to many chainsaws, and looking at the patterns they made on the Serial Plotter graph.
What I learned
We learned how to come up with back-up options when the original plan doesn't work. Majority of us were beginners in Computer Science, so we learnt how to connect the hardware with the software practically.
What's next for Tree-mendous
We will try to implement machine learning to detect sound, a better microphone so that we have fewer boxes over a larger area.
Log in or sign up for Devpost to join the conversation.