**AWARDS ISSUED PRIOR TO JANUARY 20, 2025, WERE FUNDED UNDER PREVIOUS ADMINISTRATIONS AND MAY NOT REFLECT THE PRIORITIES AND POLICIES OF THE CURRENT ADMINISTRATION.** ROOTS, A PLANT'S HIDDEN HALF, PLAY A CENTRAL ROLE IN PLANT FUNCTIONS AND INTERACTIONS WITH BIOTIC AND PHYSICAL ENVIRONMENTS. HOWEVER SIGNIFICANT GAPS EXIST FOR INCORPORATING ROOT TRAITS IN CROP IMPROVEMENT PROGRAMS. THE ROOT PHENOTYPING BOTTLENECK CONTINUES TO BE A SIGNIFICANT BARRIER NOT ONLY FOR BREEDING PROGRESS BUT TO THE FULL USE OF GENOMICS DATA. CURRENT FIELD PHENOTYPING METHODS BASED ON ROOT EXCAVATION, DIGITAL IMAGING, AND IMAGE ANALYSIS ARE CUMBERSOME AND FREQUENTLY REQUIRE SPECIALIZED EQUIPMENT OPTIMIZED FOR MODEL SYSTEMS AND WELL-CHARACTERIZED CROP SPECIES. THE MAIN GOAL OF THIS PROJECT IS TO DEVELOP A LIGHT-WEIGHT MOBILE AI PLATFORM BASED ON UNSUPERVISED DEEP NEURAL NETWORKS FOR ACCURATE AND FAST CONSTRUCTION OF 3D PLANT ROOT SYSTEMS AND TRAIT EXTRACTION THAT CAN BE USED FOR MULTIPLE CROP SPECIES. THIS NOVEL PROJECT WILL HAVE HIGH IMPACTS VIA NUMEROUS TOOLS AND ALGORITHMS, BEGINNING WITH 3D ROOT SYSTEM RECONSTRUCTION FROM IMAGES, GROWTH TRACKING, AND EXTRACTION OF KEYPHENOTYPIC VARIATIONS IN THE ROOT SYSTEM. THIS PROJECT WILL PROVIDE TOOLS THAT WILL ENABLE PLANT SCIENTISTS TO INCREASE FIELD ROOT PHENOTYPING THROUGHPUT AND LEAD TO THE PRACTICAL INCORPORATION OF ROOT TRAITS IN CROP BREEDING PROGRAMS.