From Desert to Oasis

Inspiration

There are thousands of food deserts across the United States. A food desert is a form of food insecurity. In food deserts, residents don't have access to healthy food options within a reasonable distance. This "reasonable distance" can depend on several factors like population, whether a location is in an urban, suburban, or rural area, income levels, access to transportation, price of food options, and many other things.

I want to answer an essential question: Where is the most optimal location to place a new grocery store in any given city/state to lift the most people out of food deserts?

Eventually, this project will become a full platform where grocery companies can input a zip code, city, or state and get recommendations for the best locations to open a new chain. Options (new locations) will be prioritized based on an estimated number of people who will be lifted out of a food desert.

This is a very complex question. In this hackathon, I attempted to get the ball rolling by focusing on zip codes and how they relate to food deserts.

Goal: Eliminate all food desert in the United States!

What it does

This application takes in a zip code from a user (grocery company) and tells the user whether or not that zip code is in a food desert.

A zip code is either "in a food desert" or "not in a food desert". This is determined by two google api's. Combined they calculate the distance using geolocation (longitude and latitude) of the zip code to 10-20 google maps grocery store search results. The average distance that a zip code is from a grocery store determines whether or not it is in a food desert.

How I built it

• Python, Google maps api, Google Serpapi api, blood, sweat, and tears

Challenges I ran into

I realized that much of modern research argues against using zip codes as a factor in any type of research relating to human behavior. (I didn't see this research until very late into the project)

If you look at my early commits, you will see some .html files. I gave up on the frontend once I realized that my backend was not defined well enough.

There are many factors that contribute to an area being designated as a food desert. At first, I tried including many factors into my application, then I realized that this isn't possible in the time constraint.

Finding a consensus (in the research)on what distance constitutes a food desert in urban vs non urban areas was a challenge. In the end, I created my own criteria. I opted to calculate the average distance to a grocery store. If the average distance to a grocery store for a given zip code is less than 1.5 miles (urban areas), or greater than 10 miles (non urban areas)... It's a food desert.

I understand this is a very rough definition, but for the sake of simplicity, this is good proof of concept.

Accomplishments that I'm proud of

• Finishing the pilot for this application!
• Translating my vision into a product.
• Working by myself for the first time at a hackathon.

What we learned

• 1. Learning how to research a topic in depth, and produce a tech related project that reflects what I have learned.
• sources:
• https://www.ers.usda.gov/webdocs/publications/45014/30940_err140.pdf
• https://thehumaneleague.org/article/food-desert
• https://www.socialpolicylab.org/post/grow-your-blog-community#:~:text=Food%20insecurity%20occurs%20when%20food,away%20in%20a%20rural%20area.
• https://www.aamc.org/news-insights/54-million-people-america-face-food-insecurity-during-pandemic-it-could-have-dire-consequences-their
• https://www.aecf.org/blog/exploring-americas-food-deserts
• https://www.verywellhealth.com/what-are-food-deserts-4165971
• 2. Using multiple api's in a single project. I have only ever worked with 1 api in a personal project.
• 3. Learning how to create a simple website with flask
• 4. Working across multiple python files. This was the first time I have done this on my own
• 5. I worked by myself on this project to really push myself to be independent. I definitely had to improve my googling skills. I also got really good at reading documentation.

What's next for From Desert to Oasis

• Find a better way to define what a food desert is that is based on more factors, so that a more accurate conclusion can be reached.
• Feature addition: if a zip code/ state, country ,etc, is given that is not in a food desert, recommend a location that is in a food desert. More specifically, recommend a location where the greatest number of people will be lifted out of food insecurity.
• Feature addition: add a front end component that can make everything look nice. Use bootstrap and flask.
• Feature addition: Use 2020 census data as a reliable factor in determining where a food desert is.
• Feature addition: Use matplotlib or another visualization tool to show all this data on a map that can be displayed on a website.
• Feature addition: Consider current prices for common healthy goods. How expensive are they in any given area. How does price create food deserts even when a grocery store is within a reasonable distance?

Built With

• google
• python