SmartBin

Inspiration

In third-world countries like India, recyclable and non-recyclable waste is usually misclassified and even a lot of people don't remember where their trash should go. Trash Misclassifications can even have a major impact on increasing greenhouse emissions, which further contribute to climate change.

What it does

It is a smart bin that is currently using my laptop's webcam to take an image of the material, and then it sends that image to Google Cloud. With the help of Google's CloudVision API, it identifies what is there in the image and returns it to my client. Depending upon the name of that object, my program classifies it as either trash or recyclable material and then opens up one compartment accordingly.

How we built it

In the hardware part, I've used C++ to establish control of the servo motors. In the software part, Python is used to connect to Google CloudVision and vice versa. More information is given in my readme file (Github).

Challenges we ran into

Establishing a connection between Arduino and Python was a challenge, and getting started with google cloud was a bit difficult because of transaction issues.

Accomplishments that we're proud of

Finally getting to work with Google Cloud.

What we learned

I learned about new ways to read into the serial with python and got to know more about how I can use Google Cloud in my projects.

What's next for SmartBin

Once the lockdown is over, I'll purchase and integrate it with a standalone camera and change its interface to Raspberry Pi so that it doesn't require anything external.