Our inspiration for creating CarbonAI came from our desire to help educate people on the impact their daily habits have on global warming. Our thoughts can be summarized by this quote from Amanda Little:

“Most of us generate more planet-warming emissions from eating than we do from driving or flying. Food production now accounts for about a fifth of total greenhouse gas emissions annually, which means that agriculture contributes more than any other sector, including energy and transportation, to climate change.” - Amanda Little, The Fate of Food

What it does

CarbonAI is our machine learning algorithm that we built in order to spread eco-awareness. With our user-friendly website, anyone can upload a picture of food and learn about how much carbon was emitted to make that food. All the user has to do is copy the desired image’s address, paste it when prompted, and hit submit. The program will then inform the user on how much carbon was emitted due to each ingredient.

How we built it

For our machine learning algorithm, we used Clarifai's food recognition API and implemented it using Python for our backend. We then calculated the carbon footprint based on the outputted ingredients from Clarifai, finding our data from various online sources. In order to make our frontend of the website where we take in image URLs and display the results, we used Flask and javascript. Finally, we used Heroku to host the site.

Challenges we ran into

Despite having previous coding experience with Python, none of our team members had much experience coding with front-end frameworks such as Flask. It was troublesome integrating the front-end to work seamlessly with the backend. We wanted to create a cohesive product that anyone could use without setting up the project on their computer.

Accomplishments that we're proud of

We’re proud of creating a working product anyone can use and test out right now by going to

What we learned

We learned a lot from BadgerHacks, from new technologies like Flask to what it’s like to work as a team using git for version control (the power of friendship!)

It was also really eye opening when we were doing research about climate change issues while scouring the web for data.

What's next for CarbonAI: Carbon Footprint Calculator

CarbonAI still has many features that can still be worked on. We plan on improving the UI of the website, expanding our dataset, and improving the accuracy of our model after BadgerHacks2021.

Share this project: