smartBin

Inspiration

Waste Management: Despite having bins with specific labels, people often put waste into wrong bins which lead to unnecessary plastic/recyclables in landfills.

What it does

Uses Raspberry Pi, Google vision API and our custom classifier to categorize waste and automatically sorts and puts them into right sections (Garbage, Organic, Recycle). The data collected is stored in Firebase, and showed with respective category and item label(type of waste) on a web app/console. The web app is capable of providing advanced statistics such as % recycling/compost/garbage, your carbon emissions as well as statistics on which specific items you throw out the most (water bottles, bag of chips, etc.). The classifier is capable of being modified to suit the garbage laws of different places (eg. separate recycling bins for paper and plastic).

How We built it

Raspberry pi is triggered using a distance sensor to take the photo of the inserted waste item, which is identified using Google Vision API. Once the item is identified, our classifier determines whether the item belongs in recycling, compost bin or garbage. The inbuilt hardware drops the waste item into the correct section.

Challenges We ran into

Combining IoT and AI was tough. Never used Firebase. Separation of concerns was a difficult task. Deciding the mechanics and design of the bin (we are not mechanical engineers :D).

Accomplishments that We're proud of

Combining the entire project. Staying up for 24+ hours.

What We learned

Different technologies: Firebase, IoT, Google Cloud Platform, Hardware design, Decision making, React, Prototyping, Hardware

What's next for smartBin

Improving the efficiency. Build out of better materials (3D printing, stronger servos). Improve mechanical movement. Add touch screen support to modify various parameters of the device.

Built With

• ai
• classifier
• firebase
• firestore
• gcp
• google-cloud
• hardware
• iot
• machine-learning
• pi
• python
• raspberry-pi
• react