Produce

Inspiration

Inspired by the themes of IoT and creating greener workspaces, we decided to tackle an often overlooked issue that is both costly and wasteful: excess and unused produce.

What it does

Two part application:

Track what you bought: Cashier scans QR code on each item, which pushes the item information to our database

Track what happens after you buy it: Identify item type and what items go in and out of the refrigerator

How we built it

Use a MySQL database to store item information received from QR code Use an ultrasonic sensor to activate camera inside the refrigerator Camera uses tensor-flow API for object recognition

Challenges we ran into

Inexperience with certain technologies (i.e. transmitting RaspberryPi data over a Wi-Fi network, integrating object recognition API into Python scripts for camera/sensor data)

Accomplishments that we're proud of

Able to make object identification work and identify when objects are entering/leaving the refrigerator

What we learned

Large project, with so much focus on the backend, we lacked balance on the UI front

What's next for Produce

Several metrics and features we wished we could have implemented: Notifications close to date of expiry Distinguish items from one another (i.e. milk from last week vs. milk from yesterday) UI for consumers (access product information, expiration dates, what they have used and wasted) Detach retail participation (no need for QR codes to be placed on every item, generate QR code from consumer receipts so only one party is involved)

Built With

• css3
• html5
• mysql
• php
• python
• raspberry-pi
• tensor
• ultrasonic-sensor