Project Overview: The Nourish Network is an AzureML based application designed to support food banks, mutual aid groups, and related organizations in reducing food waste and addressing food insecurity. Created for the Hack for a Sustainable Planet executive challenge, this solution analyzes historical data on food surplus and scarcity to predict potential excess, enabling better coordination for upcycling and donation. By redirecting surplus food to communities in need, the application aims to reduce waste and free up resources for other essential programs. The project was inspired by the vast 120 billion pounds of food wasted annually in the U.S., where nearly 35 million Americans, including 10 million children, face food insecurity.
Our research explored large-scale hunger relief models like Feeding America, the largest relief organization in the U.S., which connects individuals in need with local food banks. However, a gap exists in coordinating resources between donors and food banks, leaving some under-supplied. A tool like Nourish Network could streamline this, boosting donations and reducing operational costs. For instance, let's take the work of the Capital Area Food Bank in Washington, D.C., a direct partner of Feeding America. In 2023, this organization distributed 60.9 million meals, of which 31.6 million were purchased while 32.3 million came from government programs or donations from retail sources, wholesalers, and manufacturers. This significant dependence on purchased meals resulted in $95,943,433 in food costs. By improving access to surplus donations, The Nourish Network could potentially reduce these expenses, allowing food banks to redirect resources toward other critical needs and expand their reach.
Detailed Objectives: Current technology is donor-initiated, connecting donors with excess supply to food banks in real time. This limits food banks' ability to plan for future needs, forcing them to rely heavily on cash donations for food purchases. As a result, a significant amount of food still ends up in landfills.The Nourish Network shifts this model to a consumer-initiated approach, where food banks can request resources based on forecasted needs. Using AI/ML to predict food surplus, it matches food banks with donor facilities and allows for recurring donation arrangements outside the app. Insights from Feeding America and Capital Area Food Bank confirm that this approach could address critical gaps. A 5% increase in surplus donations alone could yield 1 million additional meals and save $40 million annually in D.C.