Partners: GroenvermogenNL, Utrecht University, CWI, Erasmus University, TNO, HAN University of applied Sciences, HANZE Hogeschool, Fluidwell, Nobian Industrial Chemicals, and Netbeheer Netherlands

This research project is funded by GroenvermogenNL (refer to the two links below) and we will closely collaborate with all our partners to develop AI methods and tools for Hydrogen transportation, offshore and storage. In this project, we aims to develop an automated machine learning methodology capable of learning from historical and real-time operational data of hydrogen equipment. This methodology will enable prediction of degradation and lifetime, anomaly detection, and identification of potential disruption risks. These predictions will then by utilized as inputs for a multi-objective optimization model aimed at optimizing maintenance plans for Hydrogen associated equipment, including compressors, electrolyzers, and purity measurement instruments. Hyper-parameter optimization will be used to ensure the methodology fits different types of equipment and various conditions. Furthermore, the project aims to visualize solutions in a manner that is easily interpreted by decision-makers.

Visit the GroenvermogenNL news message

Visit the NWO news message

People involved:

Robin van der Laag R
Robin van der Laag
Agnese Rizzato A
Agnese Rizzato
Yingjie Fan
Assistant Professor of Logistics Optimization
Prof. Thomas Bäck
Professor of Natural Computing