Location: Food for Health of People and the Environment Lab
Project Number: 8040-52530-001-029-S
Project Type: Non-Assistance Cooperative Agreement
Start Date: Sep 1, 2026
End Date: Aug 31, 2028
Objective:
Cooperator will assess readiness of FoodData Central (FDC) to serve as foundational data infrastructure to support precision nutrition research and artificial intelligence-driven applications. Based on the results of this assessment, Cooperator will develop strategies and tools to improve FDC's readiness to support these research priorities. Currently, FDC provides multiple data types, ranging from rigorously validated reference data to label-derived and survey-based entries, each with different analytical foundations and intended uses. This project will evaluate the extent to which FDC's current data infrastructure meets the needs of precision nutrition researchers, AI systems, and the broader user community across two complementary dimensions:
1. Precision nutrition data coverage. Will assess extent to which FDC contains the compound-level data that precision nutrition research requires, will identify critical gaps, and will develop strategies to prioritize and fill those gaps.
2. AI integration readiness. Will assess whether AI systems can effectively discover, navigate, interpret, and correctly use FDC data, and will identify and prioritize infrastructure improvements that would support emerging demand.
Approach:
USDA FoodData Central is the national public resource for food composition information, and the Food For Health of People and the Environment Laboratory (FHPEL) assumes responsibility for FDC, monitors its usage and determines its future needs for interoperability with USDA and other federal databases. Precision nutrition research increasingly requires compositional specificity, including compound forms, bioactive profiles and preparation-dependent variability, that general nutrient databases were not originally designed to provide. Further, AI-driven applications are accessing FDC through its API at rapidly increasing rates, creating demands on the database’s documentation, discoverability and data architecture that the current infrastructure was not designed to accommodate. This project integrates three complementary assessment and development activities and is designed to that findings from each inform the others. 1) Precision Nutrition Readiness Assessment. Working jointly with the Agency PI, the Cooperator will evaluate FDC’s data coverage against the compound-level specificity that precision nutrition research requires. The work includes defining the relevant data categories, selecting priority food groups for assessment in consultation with FHPEL, and conducting a systematic gap analysis across FDC’s data types. A quantitative readiness scoring methodology will be developed and validated, designed for ongoing use as the database evolves. The assessment will compare coverage across FDC’s data types and against relevant external reference databases to identify where complementary data sources could address gaps and where new analytical work would be warranted. 2) AI Integration Assessment. The Cooperator will evaluate how effectively AI systems can discover, navigate, interpret, and correctly use FDC data. The work includes benchmarking FDC’s AI readiness against peer nutrition data platforms and adjacent federal databases, and synthesizing lessons from how analogous domains have addressed AI navigation challenges. A principles-based navigation framework will be developed that provides AI systems with the contextual guidance needed to use FDC data appropriately, including selecting the correct data type, interpreting quality signals, and handling incomplete data. The framework will be tested empirically by evaluating AI system performance before and after its application, producing quantitative evidence of its value. A prioritized integration road map will identify what improvements are needed, with technical implementation details deferred to engineering assessment. 3) User Segmentation Evaluation. The project will incorporate input from FDC's user communities, including researchers, application developers, food industry users and agriculture stakeholders to inform prioritization of data development and infrastructure improvements. Through interviews and facilitated discussions, both parties will contribute to this facet of the project. The agency will engage agriculture groups and food industry stakeholders, and Cooperator will engage research and technology user communities.