AI and machine learning

Artificial Intelligence (AI), notably Machine Learning (ML), has gained significant attraction across various fields, including geosciences. Accessibility to high-performance computing power at lower costs, makes applying ML techniques in geosciences more feasible. The fusion of AI not only enhances the efficiency of geoscience workflows but also accelerates processes significantly.

The researchers within the Machine Learning Group at the Geosciences Institute of Jena University are dedicated to exploring various applications, including:

•  Subsurface gas storage (H₂, CO₂) for energy transition
•  Geothermal applications
•  Pollutant transport in environmental contexts
•  Reactive transport in porous media
•  Colloid and nanoparticle transport
•  Bacterial and bio-colloid transport, and
•  Conventional oil and gas production.

Our primary focus encompasses a range of activities by AI and machine learning, including but not limited to:

•  Rock segmentation (covering binarization or multi-mineral segmentation)
•  Estimation of porous rock properties (such as permeability, flow field)
•  Resolution enhancement of porous media images
•  Image reconstruction of rocks/soils
•  Simulation speed-up
•  Upscaling from pore- to core-scale

The combination of availability of advanced multi-scale imaging tools in the Geoscience institute such as the Zeiss Xradia 620 Versa X-ray microscope in the Applied Geology Department, and  Ultra Plus SEM-EDX de (Zeiss Co.) in the Hydrogeology Department alongside with access to high-performance computing (HPC) infrastructure (Draco  deand ARA de supercomputers with parallel GPU computing capability), distinguishes our institute in the application of AI tools for geoscience applications at the pore scale.

Associated: Prof. Dr. Thorsten  Schäfer, Dr. Saeid Sadeghnejad, Dr. Sarah Hupfer, Ariunzaya Löwe, Anna Kogiomtzidis

Projects: RENA, TRAVARIS

Completed projects: KOLLORADO-e3, TransLARA, CONCERT_CCair