A complex network of stakeholders, resources, activities, and information move food from farm to plate through a soil to soil food system (FS). To ensure the health of this system, especially through times of crisis and instability, all of the system's stakeholders (e.g., residents, food-producers, researchers, and policy-makers) need fresh, up-to-date information, something only possible at scale through the aid of community members. This project lays the foundations for a sustainable, community-driven data collection tool that will simultaneously help community members to better understand their food system by answering questions like "Where can I find healthy, fresh, low-cost food?" while encouraging participation in the food system by helping to keep data fresh and up-to-date. This planning proposal will develop a proof-of-concept, using image recognition (e.g., at retail stores or farm stands) to streamline the process of recording information about the availability, price, types, and other characteristics of produce available at produce displays. The project will also work with community members to drive adoption of the tool through healthy eating workshops and data-collection competitions.

Working collaboratively with non-profit and local government community partners in the Western New York (WNY) food system, the project team of researchers and stakeholder will develop data collection platform called REAP (Research, Education, Action, And Policy for Resilience) to allow crowdsourced data to augment an existing WNY food system dashboard. For example, by allowing for crowdsourced produce price data, the team can analyze how this information impacts the behaviors of both consumers and suppliers in the food system. The project will advance the state of the art in computer vision and uncertain data management by developing techniques to correlate multiple sources of information in images while ignoring distractors, and for tracking the system's confidence in the resulting information through the entire data presentation pipeline. The proposed work also includes a meta-analysis, using pre- and post- surveys, as well as graph analysis on the data captured through REAP, to examine how the availability and flow of co-produced information and food-related behavioral actions affect food system stakeholders. REAP lays the groundwork for ongoing efforts to map the WNY food system, simultaneously studying both urban and rural participants.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.