Data Driven Discoveries for Agricultural Innovation (D3AI) - a Plant Breeder's Perspective.

Data Driven Discoveries for Agricultural Innovation (D3AI) at Iowa State University is a multi-disciplinary initiative spanning several colleges and departments to make significant strides in the collection, management, interpretation, and use of data related to agriculture at multiple scales (plants to geographical regions). This initiative has spurred several new projects that interplay between inter- and trans-disciplinary approaches for complex research topics in plant breeding and other disciplines. The plant breeding projects leverage advances in phenotyping, robotics, phenomics, data curation, and data analytics, particularly by machine- and deep-learning algorithms, and their integration in a data aware decision-driven pipeline to intensify genetic enhancement for multiple traits. These projects work on several layers of phenotyping, from rare and microscopic objects to large scale aerial images of hundreds of breeding accessions, and include physiological, morphological, performance, and stress traits. The objectives of these approaches are identification and validation of important genomic regions, trait selection, and prediction. These projects are complemented by advances in robotics, particularly for phenotyping, at previously unimaginable spatial and temporal scales. While plant breeding is, and will always remain, a data-centered discipline, we are now entering big data V (volume, velocity, variety, veracity, and value) which require data awareness and seamless pipelines from data collection, quality control, storage, analysis, reporting, and decision-making. As we enter this transformative stage of innovation, plant-breeding education will also need to adapt to this changing and evolving discipline and requires a tighter collaboration with numerous disciplines and communities.