CHARACTERIZATION OF 3D PRINTED PROTEIN INKS WITH CUSTOMIZATION USING MACHINE LEARNING

Our long-term goal is to deliver highly nutritious sustainable foods on demand through 3D printing that are optimized for manufacturability, sensory appeal, and nutrition. Our goal for this proposal is to characterize protein inks including OC/HOCs for their rheological and sensory properties, and demonstrate the feasibility for machine learning to deliver personalized printed products based on characterized relationships.We will pursue the proposal goal with the following specific objectives:Specific Objective 1: Printability characterization for protein inks. Our working hypothesis is the printability and texture of protein inks is controllable by altering ink composition. The hypothesis will be tested by creation of diverse protein inks with rheological property characterization and computer assisted assessment of printed shape fidelity.Specific Objective 2: Machine learning for personalized manufacturing. Our working hypothesis is machine learning can model complex relations among food ingredients, printability, and sensory properties in a predictable manner for customization. The hypothesis will be tested by measuring relevant properties and validating how ML enables 3DFP personalization.