top of page
1634883737564.jpg
FFG.png

AI research

sfg-logo-mit-claim-rgb-300-transparent-dpi-png.webp

GeoDrone (cofinanced by the FFG)

Brief project description:

 

The aims of the project are the development, selection and application of artificial intelligence (AI) algorithms or the new development of AI architectures for the delimitation of geological facies areas from 3D drone photos. In addition, the algorithm should be used to automatically interpret and estimate petrophysical or geomechanical properties. Users receive automatic interpretation of outcrops. So far, for interpretations, users have had to segment and name the outcrops manually, at least partially, partly without drone photos, partly with drone photos, some also with simple algorithms. An interpretation regarding faults, geomechanics and petrophysics is a novelty. With the tool to be developed, it is possible to automatically interpret the outcrops and display them three-dimensionally. Furthermore, if many similar information is available, there is the possibility that the algorithms can be trained better and better and, thus, work even more precisely and faster. The Montanuniversität Leoben with the Zentrum am Berg (ZaB) is also on board as a research partner.

Economic importance/exploitation potential:

 

The project is of great importance because the tool can be used to segment and classify an outcrop in terms of facies, geomechanics and petrophysics faster, more objectively and more precisely. Any company involved in any way with geological exploration will be interested in such a tool, including energy suppliers, mining companies and construction companies, other geological or geotechnical engineering offices and public clients.

AI for Green Wells  (co-financed by the SFG)

Brief project description:

 

The assessment of geothermal zones in boreholes regarding porosity, permeability, clay content and temperature are currently being solved using complex petrophysics. Here, geological borehole logs are interpreted in connection with measurements on cores. Johannes Amtmann already has many years of experience (approx. 15 years) in the classic interpretation of borehole logs. This interpretation will continue to be relevant, however, it is often stretched to its limits, especially with many different boreholes, but also with boreholes that differ in terms of many different borehole measurements. In addition, a tool would also be interesting to carry out the interpretations much faster to obtain initial information about the geothermal deposit. Many wells that already exist can be checked to see whether they are geothermally interesting or not. The development of the tool is of course relatively time-consuming and would not be financeable without funding; but in the long term, it offers enormous growth potential.

Economic importance/utilization potential:

 

As a result, a workflow is available: AI workflows for evaluating a geothermal deposit without using petrophysical formulas as well as the verification of an AI workflow for borehole data. The results will show the potential of the workflow. After completion, the project can be presented to the client and the advantages and disadvantages can also be explained. The aim is of course to get follow-up projects, especially from energy suppliers (Energie Steiermark, Wien Energie, Energie Linz AG etc.). Hence, the potential may be recognized, and the workflow can be further developed and expanded for their questions.

bottom of page