Fruit Price Adjustor

• 

  web visual of price

Inspiration

Wasted food has tremendous environmental and economic impacts. It accounts for about 4% of greenhouse gas emissions and 2% of U.S. gross domestic product. EPA announced a goal to reduce food waste across the United States by 50 percent by 2030. Our vision is to use this application towards reducing food waste in supermarkets.

What it does

It sets an appropriate price based on the ripeness of a fruit in order to reduce the sell ripening fruits quickly, thus reducing waste.

How we built it

A camera was connected to a Raspberry Pi. The Pi takes images every 5 seconds and sends the image to a server. The server runs a Neural Network model on the image to estimate the ripeness of the fruit. Based on the estimate, an appropriate price is set. An Express server exposes an endpoint for the ReactJS front end to get the price every 5 seconds and display the price.

Challenges we ran into

Transferring data from the backend to the frontend every 5 seconds was a little challenging. We finally solved with using Express to make an API endpoint and make requests every 5 seconds.

Accomplishments that we're proud of

We made a complete working model of an application in span of less than 2 days.

What we learned

We got familiar with integrating a Machine Learning model with a full-stack application.

What's next for Fruit Price Adjustor

Add more models to expand to other types of fruit. Fine tune the machine learning models to more accurately detect the desired level of ripeness.

Built With

• express.js
• imagezmq
• node.js
• python
• react
• tensorflow
• twilio