Inspiration
Food deserts are areas that experience a combination of low-income and inaccessibility to grocery stores providing fresh, healthy food options, and they're more common than people might expect. This project is intended to raise awareness for the factors contributing towards making a food desert and how their own city is impacted.
What it does
A user inputs their zip code and receives a quantitative score of the food desert severity in their area, calculated with a heuristic that takes median income, distance to grocery stores, vehicle access, and various other low-income and low-access indicators.
How we built it
This project was built using the USDA Food Access Research Atlas dataset, Gemini, and the HUD-USPS ZIP Code Crosswalk dataset. It uses Flask for the backend and HTML/CSS for the frontend. The datasets: https://www.ers.usda.gov/data-products/food-access-research-atlas https://www.huduser.gov/portal/datasets/usps_crosswalk.html
Challenges we ran into
I am only one person, so I wasn't able to accomplish everything I initially planned for the project. The census data needed cleaning and didn't cleanly map to specific, intuitive locations like cities or addresses, so I needed to use the separate HUD-USPS API to convert zip codes to the census tracts. I also hit the Gemini API usage limit and could not implement the Gemini features I wanted to.
Accomplishments that we're proud of
I'm proud that in the end, I was able to work my way through countless errors and bugs and make a working product that computes a fairly intuitive score indicating how much of a food desert a location is.
What we learned
I got to work with datasets I hadn't known about previously, and got more comfortable with building custom API endpoints and backend web development.
What's next for Food Desert Locator
I originally intended to utilize the Gemini API to analyze the specific data for the given location and generate a customized explanation of all the factors contributing to the score and potential solutions for accessing fresh groceries. I also planned to make use of Google Maps data with Gemini to locate nearby grocery stores and other resources.
Log in or sign up for Devpost to join the conversation.