Our approach
iML-WKW builds the shared foundation that organizations need to develop AI models for weather, climate, and water in a reliable and reproducible manner. This ensures that every organization does not have to reinvent the wheel individually.
To achieve this, the facility focuses on aspects that recur in almost every research project: standardization of work processes, reproducibility of results, access to computing power, and systematic quality assurance. By addressing these challenges collectively, iML-WKW acts as a catalyst for AI applications in the urgent weather, climate, and water issues facing the Netherlands.
Guiding Principles
The facility has been developed based on a number of guiding principles that apply to all partners and users.
Open, unless otherwise required
code and data are made public by default.
FAIR
data and models are findable, accessible, interoperable, and reusable.
Explainability and traceability
every result is reproducible and traceable.
Modular and scalable
new models and data sources are easy to integrate.
Agile methodology
iterative building, from prototype to production.
Collaboration and open access
iML-WKW operates through an open innovation infrastructure. Universities, water boards, engineering firms, and private sector partners can join and contribute to the development of modules. Use cases and users are actively facilitated through the DigiShape network.
The collaboration is aimed at rapidly moving promising technologies from proof-of-concept to practical application, without each party needing to make large individual investments.