Project Reducing Food Waste

Visualization of contributors of food waste highlight the impact of Grocery Stores and Supermarkets
This image has the data frame we worked with to create the map visualization.

Inspiration

Some of our team members have worked in the food industry, and it's known that a significant amount of food goes to waste due to inventory issues.

What it does

After adding all the information about the products, you will be alerted if a product is about to expire and what aisle the soon-to-expire product is located in. The code gives you the opportunity to put a product on sale and it will internally update the price of the item. If you choose not to put the product on sale, it will recommend next steps for what to do with the soon-to-expire product.

How we built it

Using a dictionary of dictionaries to create a O(1) data management system that is easy to edit and update.

Challenges we ran into

When analyzing the data sets, we encountered difficulties in writing a function to classify the names of the grocery stores using Google Maps and long and lat values. To solve this issue, we manipulated the data frame with the longitude values so that they would appear correctly on ArgGis.

When initially creating the architecture of our system, we wanted to organize the inventory data into nested lists, but soon realized that was complicated and had a O(n^2) run time, which would compromise the efficiency of our inventory management system. After deliberation, we switched to our current dictionary system.

Accomplishments that we're proud of

On the data science side, we are proud to have used a raw data set, cleaning it by splicing rows and isolating specific classifications. We were able to filter through many layers of data and determine that Stop & Shop is the highest food waste producer in Massachusetts.

What we learned

We strengthened our skills in creating and manipulating data in Python, as well as being creative in making our data visualizations.

What's next for Project Reducing Food Waste

With more time, we will specialize our program to accommodate all shapes and sizes of a grocery or supermarket. Our program will also eventually include a feature to systematically determine whether food should go to a food bank/shelter or remain in the store's inventory.

Built With

chatgpt
esri
excel
googlecolab
numpy
pandas
python
rstudios
sklearn
spyder
vscode

Submitted to

DS+X
- Winner FirstHack - 3rd

Created by

At this Hackathon, I performed ETL processes on public Massachusetts databases, cleaning and merging disparate datasets for grocery store waste analysis;
utilized Python (Pandas, NumPy) to engineer features and manipulate variables, creating data visualizations that highlighted inefficiencies in corporate waste management &
conducted geospatial analysis with ArcGIS to identify and define high-density geographic clusters of food waste across corporate grocery chains. My work using ArcGIS inspired me to further pursue projects and roles that use GIS tech!

Sameeha Shaik
I conducted research then turned an EPA table in a csv and developed data visualizations in R to demonstrate the severity of retail food waste and model our application's potential impact.

Adedayo Sanusi
My main contributions to our hackathon project were in data preparation and presentation. I handled much of the data cleaning process to ensure the dataset was accurate and ready for analysis and visualization. I also played a key role in shaping how we presented our results, both through slide design and explaining our findings. In addition, I helped think through the front-end design of our project to ensure it was consistent with our idea and a product grocers would actually use!

Allison Lee