The objective of this project was to investigate the feasibility to apply the Farm Data Train concept in the domain of Wageningen University and Research. The Farm Data Train is based on the Personal Health Train, which is an infrastructure designed to bring algorithms to the source of the data instead of bringing the data to the algortihms. This ensures the protection of the patient’s privacy as the data never leaves their hospitals. As a prerequisite, the data used in the train should be made FAIR: Findable, Accessible, Interoperable and Reusable. At Wageningen University and Research issues involving the privacy and sensitivity of research data are also encountered. The goal of setting up a FAIR Data Train in Wageningen ensures safe, efficient and ‘in-control’ sharing and re(use) of data to be able to better address data-driven research questions while respecting data ownership and data privacy regulations.
The core elements of a data train are the data stations, the train, the data gateway and the tracking system. The data stations provide FAIR access to data and metadata by allowing a train access to interact with the data. The trains are the models that process the data and are able to learn from it. The data gateways take care of the authorization and authentication of the trains. Lastly, the tracking system manages the transport infrastructure and the routing of the trains.
The goal of the project is to develop a prototype data train between two FAIR data stations in Wageningen to show the proof of principle and to identify any knowledge gaps.