At OpenGeoHub, Robert supports current and future high profile European Commission-funded and other international projects where there is a need to develop new solutions for geocomputing, optimizing modeling frameworks and publishing the scientific outputs. Specific task include but are not limited to developing code and datasets that aim at:
- Helping with preparing, harmonizing and cleaning point datasets and soil spectroscopy datasets that can be used for training predictive models,
- Running predictive soil mapping using Machine Learning in modeling frameworks,
- Implementing parallelisation in cluster or cloud solutions,
- Using HPC-C4G to optimize spatial analysis increasing mapping accuracy of automated predictive mapping.
Robert holds PhD in Physical Geography and has a background in Geoinformatics. He has experience with digital soil mapping and predictive modeling of forest disturbance using machine learning. At OpenGeoHub is a postdoctoral researcher. Besides the above-mentioned tasks, his function also includes the publication of results in peer-reviewed journals and supervision of junior staff.