AI-FEED (Healthy Food Access)

Mapping Healthy Food Access & Suggesting AI-Driven Solutions

INTRODUCTION

Food insecurity continues to pose significant challenges in urban environments like Greater Houston, where factors such as uneven economic development, limited healthy retail options, and inadequate transportation converge to restrict access to nutritious foods. In response, our team has developed the Healthy Food Access Index (HFAI), a multidimensional measure capturing economic and sociocultural disadvantages, physical accessibility, and the local retail environment. We couple traditional GIS-based mapping with AI-driven strategies, culminating in our “AI-FEED” web platform. AI-FEED equips community leaders and urban planners with a robust toolset for pinpointing critical food deserts and implementing tailored interventions aimed at improving overall food security.

METHODOLOGY & HFAI DEVELOPMENT

The Healthy Food Access Index (HFAI) integrates five sub-indices: Economic Deprivation (EDI), Sociocultural Deprivation (SDI), Retail Food Environment (RFEI), Physical Access (PAI), and Accommodation (AAI). Each index draws on data from reliable sources—such as the American Community Survey, USDA, and WIC store locators—to evaluate community-level factors influencing food choices. Weighted and standardized, these indices form the composite HFAI, spotlighting regions of Greater Houston where affordability, geographic reach, and the cultural context of food are most lacking. This holistic approach acknowledges that “distance to a store” only scratches the surface: family income, neighborhood culture, and local transit options can all dictate how easily healthy foods are accessed and used.

RESULTS & HOT SPOT ANALYSIS

Leveraging the Getis-Ord Gi* statistic in ArcGIS, our analysis revealed distinct “hot spots” (areas with notably low healthy food access) and “cold spots” (areas with comparatively better food environments). Neighborhoods in East and Northeast Houston consistently surfaced as hot spots, whereas more affluent zones—particularly in West Houston—fared better with greater supermarket density and improved infrastructure. These results confirm that limited access often ties back to public transit gaps, socioeconomic hurdles, and an inconsistent distribution of healthy retail outlets.

AI-FEED: WEB-BASED PLATFORM

This analysis led to our creation of AI-FEED—a dynamic web platform that translates our mapping and indexing efforts into actionable insights. Beyond visualizing essential data like neighborhood demographics, obesity or diabetes rates, and store locations, AI-FEED can propose real-time recommendations. For instance, it may identify prime neighborhoods for new bus routes, or suggest where existing corner stores could stock healthier items to boost community nutrition. Planners can also examine how new food pantries, incentives at dollar stores, or community gardens might alleviate local food insecurity. Together, these features allow city officials and community nonprofits to transition from awareness to impactful, data-driven action with a simple dashboard.

DISCUSSION & FUTURE WORK

The HFAI framework and our subsequent hot spot analysis highlight the intricate social, economic, and infrastructural factors driving food deserts. Next steps involve integrating streaming data sources—potentially connecting AI-FEED to large language models that monitor local news or community forums for sudden store closures or new development initiatives. AI-driven scenario planning could also be extended to evaluate the long-term impact of improved bus routes or targeted pantry expansions. Looking ahead, we envision bringing these real-time, data-rich insights to other major cities across the United States, and eventually applying these insights on a global scale to help communities tackle food insecurity head-on.