Non-invasive phenotyping with autonomous robots

Recent technological advancements in non-invasive sensor systems have revolutionized field studies in plant phenotyping. Proximal sensing technologies such as the Fluorescence box (FLOX) or the light-induced fluorescence transient (LIFT) alongside with remote sensing methods from satellites (FLEX), aircrafts (HyPlant) or UAVs enable non-destructive measurements under realistic environmental conditions on different scales. The expanded deployment of these technologies, coupled with the integration of automated systems or machine learning algorithms aiding in measuring and analyzing data respectively lead to a massive increase in both the volume and complexity of data. In response to this data surge, it is imperative to implement sophisticated data management strategies that enhance the findability, accessibility, interoperability, and reusability (FAIR) of data. Such strategies are essential not only for facilitating access by human researchers, breeders and other stakeholders but also for improving machine readability and enabling effective communication among automated systems. 

To achieve this objective, we are developing a comprehensive benchmark dataset that exemplifies typical field phenotypic and remote sensing data. This effort is supported by advancements in research data management, including templates for data organization. Additionally, we are committed to enhancing FAIRness through the creation of guidelines designed to educate researchers and data publishers on effective data documentation, structuring, and licensing practices. By establishing these standards and contributing to FAIRagro, we aim to significantly enhance the reproducibility and utility of field phenotyping data for the whole community.