AQUABIN
AI , IOT and Remote Access enabled system to Detect , Maneuver and Collect floating trash over
the water bodies using Computer Vision and self motion planning with monitoring report generating system connected
with easy to use game like UI.
According to the ocean index report 260 aquatic species die because of water pollution in the world and about 30% of 80% of 1.5 metric tonne of open waste in India is left in the water bodies according to india today report which become a threat to our nature.
And affects the life of animals, humans , tourism and the economy of our country. And it is very difficult to clean the floating trash manually from small urban drains, streams, rivers, lakes and estuaries , and have life hazards and lack of proper technology.
To overcome this problem we come with project AQUABIN which is a IoT and AI Expert System compatible with any Trash collecting boat which can be configured with the minor changes and need almost no human intervention on the site.
We found that most of the focus in this field is on design of the physical body of the boat and boat operator is always required onsite , we found robotic boats but they are expensive as they use GPS, sensors , Floating Check Points , Complete Analysis of the water bodies and lags in Motion Planning and trash detection which reduce their accuracy and speed.
To overcome the problem of motion planning and to make a system at low cost and with better accuracy we developed the Expert system, which does this task automatically with computer vision and self motion panning
Due to lockdown we are not able to build the proper body of the boat so we tested our model on the wheels , we developed the dummy car with the help of wooden body, webcam, arduino boat ,DC gear motor and other hardware , to test our IoT control system and android application both connected via cloud API
For AI code we developed an simulator to test the working .
System works under 2 modes - Manual and AI
1. Manual
Under the manual mode your app will get connected to the cloud API and user authentication ,authorization and validation will done.
After that you will get a live stream from the boat camera to your app dashboard. So that you can control your boat in real time to detect and collect the trash. Once the bucket gets full or battery status gets low you will get a notification to charge the boat or to empty the trash bucket
. If you are at a remote location and internet connection is not available you can connect boat over wifi module also but due to low bandwidth you will not get a live stream from the boat.
2. AI
In AI mode , the boat will start detecting the trash via Computer Vision and self motion planning and give instruction to the boat controller and keep the track of each movement and display it in form of monitoring records. Modes of AI depend upon the water bodies.
You can see working of the boat in a 2D top view simulator.
