The rapid development of plant-based foods has provided opportunities for growers to explore new strategies for agricultural production with positive ecological and economic impacts

The Department of Agriculture National Institute of Food and Agriculture awarded North Carolina Agricultural and Technical State University $750,000 on December 15, 2025, under the 1890 Institution Capacity Building Grants program (CFDA 10.216) to develop an artificial intelligence-driven production system for converting grain peas into pea protein isolate and to integrate foundational and applied AI teaching modules into the university's food science curriculum.

The project will optimize pea protein extraction through machine learning-based ultrasonic extraction and pH-shifting combined with sonication treatment to modify functional properties of pea protein isolate. A second objective uses deep reinforcement learning algorithms and digital twin technology to optimize non-thermal pea protein drying for energy reduction and quality improvement. Work performed at Greensboro, North Carolina will establish real-time sequential decision-making capabilities to reduce processing costs and enhance product quality, creating additional income streams for North Carolina grain pea growers currently producing feed-grade commodity.

NC A&T will develop AI teaching modules covering foundational AI concepts integrated into food chemistry, food microbiology, and food processing courses, with case studies demonstrating real-world application through the pea protein optimization work. The project will train next-generation food technologists with enhanced AI capacity in support of workforce development objectives for historically black colleges and universities. A computer scientist and computational engineer at the University of Illinois will collaborate on curriculum development and optimization research. Performance extends through December 14, 2028.

The main goal of this integrated education (EP) and research (EQ) project is to establish an AI-driven production modality to convert grain peas currently used primarily as an animal feed in north carolina into PEA protein isolate thereby creating and providing an additional income stream for the growers. Additionally, the project seeks to significantly enhance the food science educational program at NC a & t by integrating both foundational AI and applied AI teaching modules into the core curriculum. The specific objectives are 1) enabling AI-assisted PEA protein extraction and improvement of the functional properties of PEA protein including machine learning-based optimization of ultrasonic extraction of PEA protein and PH-shifting and sonication combined treatment to modify functional properties of PEA protein isolate; 2) AI-enabled optimization of non-thermal PEA protein drying process for energy reduction and quality improvement achieving a dynamic real-time sequential decision-making optimization empowered by deep reinforcement learning algorithms and digital twin; and 3) training of next-generation food technologists with enhanced AI capacity. By working with a computer scientist and NC a & t and a computational engineer at university of illinois we will develop teaching modules covering foundational AI concepts in food chemistry food microbiology and food processing courses as well as to create case studies that demonstrate real-world application of AI utilizing the work related to optimization of PEA protein production. The major outcome of this project would be an increased number of skilled workforces that support executive order for excellence and innovation at hbcu.