Application of artificial intelligence in the geosciences

Country

South Africa

Sector

AI research.

Project Aim

The CGS is investigating
the possibility of applying AI to solve complex geoscientific problems.

Application of artificial intelligence in the geosciences

The application of artificial intelligence (AI) is currently receiving major interest from many sectors of society. From its origins in the 1950s, the field has seen steady progress made in the past 50–60 years in basic AI research. The availability of significant amounts of data and vast advances in computing power have recently resulted in accelerated growth in the field. The CGS is investigating the possibility of applying AI to solve complex geoscientific problems. AI systems are powerful and allow for the automation of many processes. In addition, AI is excellent in identifying patterns and relationships in large volumes of data, far exceeding human efforts unassisted by such technology. Thus, the application of AI in geoscience projects involves the design of systems that will help scientists to increase their work efficiency and help to solve some complex geoscience problems.

Despite the widespread application of AI, the CGS selected the following areas, based on their social relevance and perceived impacts and their potential to be scaled to a national level:

AI in groundwater studies — this phase of the project intends to address the prediction of mine water quality in a mining area. The information is critical to the development of passive treatment plants. The phase also involves the development of tools for groundwater potential mapping.
AI in dolomite studies — development of an automatic borehole classification system and a pre-warning system for sinkhole and related subsidence.
AI in mine seismology studies — this phase involves the development of routines to enhance our understanding of mining seismicity in support of health and safety in mines, as advocated by the Mine Health and Safety Council.

The main aim of the research is institutional capacity development in the application of AI in the geosciences. The overall research objectives are:

Developing software packages that can be used to efficiently extract information and knowledge from massive volumes of geoscience data (mine seismology, hydrogeology, borehole, etc.).
Developing a prototype of a pre-warning system for sinkhole and related subsidence.
Developing the capacity of scientists in the application of AI in various geoscientific fields.

The significance and intended contribution of the project to geoscience knowledge and to the people of South Africa are as follows:

The project seeks to push the limits of scientific understanding in regard to building early warning and prediction systems.
The project seeks to come up with better ways to extract valuable information from geoscientific datasets.
The research will contribute the following to the people of South Africa:

Efficient extraction of knowledge from geoscience datasets and utilisation of the derived products such as groundwater and mineral potential maps to address socio-economic development initiatives for the areas studied.
The development of a pre-warning system will go a long way in pre-warning the relevant authorities in South Africa to evacuate certain areas to give them ample time to act before infrastructural collapse. In the process, it is likely that lives will be saved.