IBM has been working with SIRIUS to identify opportunities for collaboration around use cases in both oil and gas exploration and production, using technologies like artificial intelligence, optimization, and cloud-based platforms. As a case in point, we look forward be being able to train our systems on relevant real-life business data. We value being able to get a variety of viewpoints on how to solve specific challenges by working with other researchers and partners.
In this collaboration, IBM has engaged with our Research division, in particular the Yorktown Research Centre in the United States and the Brazil Research Centre in Rio de Janeiro and São Paulo.
IBM Research Brazil, led by Ulisses Mello, focuses on research related to oil and gas. This aligns very well with SIRIUS´ aims. For example, the University of Oslo, Equinor and IBM are working on a project to develop tools to annotate geoscience images. Images must be annotated to train machine learning tools to allow geoscience images to be more precisely classified. Retrieving and processing information from large databases of geological documents is challenging, especially as lots of images are embedded in the documents. In this way, we want to create better AI-based tools for automatically detecting, segmenting and maybe characterizing geological images in documents. Good annotation needs annotators with high levels of geological knowledge and skill. Our aim is to support annotators with a tool that allows quick, correct and consistent annotation of images, in the first instance, geological maps.