Awards issued prior to january 20, 2025, were funded under previous administrations and may not reflect the priorities and policies of the current administration.** a major bottleneck in our ability…

Awards issued prior to january 20, 2025, were funded under previous administrations and may not reflect the priorities and policies of the current administration.** a major bottleneck in our ability to improve crops is a GAP in our understanding of how genetics and environment interact to produce plant phenotypes. Plant phenomics aims to reduce this bottleneck by developing imaging/remote sensing and computational approaches to non-destructively measure and model plant physical and physiological properties in high throughput and at high temporal resolution. This is an inherently technology-driven approach, so the plant phenomics community draws on expertise from the biological/agricultural sciences, computer and data sciences, engineering, statistics, and mathematics to solve problems in data acquisition, data management and computing infrastructure, and data analysis and interpretation. While extraction of information from massive amounts of image and sensor data is still a growing area of research in plant and agricultural sciences, an abundance of algorithms for information extraction from these data types exists from the computer science, geospatial, medical, and other communities. In particular, the availability of image and other datasets on the internet has led to a surge in methods to mine information from these massive datasets. However, while these algorithms can be leveraged for plant phenotyping problems, several common problems with implementing them in agricultural research exist. First, most image analysis tools and algorithms were designed or trained to work with images of people, cars, buildings, and other anthropocentric scenery. Second, these algorithms are implemented separately in a variety of programming languages and platforms that make IT more difficult for a non-specialist to use for another purpose or to integrate with other methods (interoperability). Finally, new algorithms are produced at such a rapid pace that IT is difficult to discern the state-of-the-art or make decisions about which methods to TRY at any given point in time. Through this alliance of plant phenotyping software (apps) developerscoordinated innovation network, we hope to coordinate efforts by: 1) bringing together developers of existing plant phenotyping software to make tools interoperable; 2) bringing together a multidisciplinary group of researchers to develop algorithms that will extract additional information from existing agricultural data and develop tools that tackle current bottlenecks in agricultural data analysis; 3) integrating existing 'stand-alone' tools/algorithms for plant phenotyping into sustainable open-source tools for the community; 4) establishing a sustainable network of phenotyping software developers and users that promotes reproducible research and reusable and extensible software; and 5) engaging phenomics software users and developers from the public and private sector through hands-on workshops, tutorials, and network events and activities. To be clear, our aim is not to build a single monolithic software package that consumes all others. Rather, this network will integrate tools where IT makes sense and will connect tools when IT makes sense to keep them as separate entities. Ultimately, we want the network to build an ecosystem that incentivizes software tools that enhance reproducible research and open publishing of code and datasets.