SIRIUS - Centre for Scalable Data Access

The SIRIUS Centre for Scalable Data Access aims to enable the digitalization of knowledge-intensive industrial work processes. Here, access to data is a bottleneck in obtaining business benefits. Unfortunately, the information used in industry is hard to access. It is spread over many databases and has inconsistent formats. Transfer of information, even within a company, is hard and time-consuming. Costs increase and poor decisions are made. This was identified as a problem for national competitiveness in a KONKRAFT report in 2018. SIRIUS brings academics, IT vendors and industrial end-users together to address this problem, with focus on the oil & gas value chain. We brought together operators with their suppliers to identify business challenges. We worked with IT vendors to develop solutions to the industrial challenges. From its initial focus on oil & gas, we have followed industrial trends and have developed results that have broad application and value. Digitalization is about using computers to change how work is done. SIRIUS therefore had a double innovation agenda. First, we created IT innovation through new hardware and software. These offer better ways of working, better data handling and easier access to information. This builds on our research programs, where we bring together disciplines in computer science to develop methods and tools. Our ambition was to combine results from these disciplines in practical applications in industry. Digitalization requires business solution innovation. Here IT tools are used to transform work processes. SIRIUS has addressed this through beacon programs for subsurface, engineering and operations and cross-domain applications in health and environmental studies. Research-driven IT and business innovation is challenging. SIRIUS has analysed the interplay between technology and organization and contributed to the NORSOK strategy for technical information requirements. SIRIUS’ innovation is built on informatics. Semantic technologies are methods that allow information to be created by giving meaning to data. SIRIUS has advanced this discipline by developing OTTR, a tool that allows engineers to create semantic models themselves. SIRIUS also developed the Information Modelling Framework (IMF) for modelling the information needed to design and run industrial assets. Our database research has improved the performance RDFOx, a graph database developed and commercialized by our partner at the University of Oxford. Formal methods develop mathematical tools used to analyse complex systems. The aim of this work is to simulate systems and verify that they work properly. The SIRIUS team has further developed ABS, a software system that forms the basis for several EU projects. Our domain adapted data science work has looked at combining technical information in documents with structured data and knowledge with machine learning. This improves the performance and explainability of the machine learning. This is one of the pillars of Integreat, a new Centre of Excellence (SFF). Data access and decision making require fast and efficient calculations. SIRIUS’ HPC and cloud computing researchers have worked on improving the performance of simulations of oil & gas reservoirs. We have also contributed to improved HPC architectures and high-capacity switches. The research in this program has initiated several EU projects. SIRIUS has demonstrated business innovation through its beacon programs. In the subsurface data beacon, we have demonstrated a Geological assistant, a program that supports a geologist in documenting and analysing their reasoning. We also demonstrated how knowledge-based data access can improve how end-users extract data from databases, themselves, without IT support. This is demonstrated on published subsurface data from Equinor’s Volve field. The facility engineering and operations beacon has looked at business problems related to the engineering, construction, and operation of complex facilities (like oil platforms). Here we worked together with a series of Joint Industry Projects. We have worked with engineering firms and operators to explore how IMF can be used to model facility information. This has led to a promising engagement with international standards and German industry. OTTR is used as a tool for building the IMF models. The digital twin is a pattern for building applications that solve real business problems. SIRIUS has developed best practices and architectures for digital twins, in collaboration with Brazilian partners. The ABS tool can be used to simulate and analyse logistics and plans. The approach works for oil & gas supply boats, it can also be applied to planning in hospitals. SIRIUS has applied complex computer science to business challenges in the oil & gas value chain. We see now that the twin transition – digital and green – demands more of our information management systems. SIRIUS has provided a foundation for this transformation.

Actual impacts: - IMF as an promising cross-industry framework for transforming information management in engineering and operations. - OTTR as a tool for making semantic modelling scalable and usable in real-world applications. - The SIRIUS digital twin model and best practices as a pragmatic tool for understanding implementing cost-effective, scalable digital twins. - Establishment in European and global networks for industrial application of semantic technologies. - Supporting and enabling a new wave of standardization and interoperability in the process, energy and manufacturing industries: DEXPI, NAMUR, IDO, IDTA, CFIHOS/IOGP, OSDU. - A better, faster and more accurate open-source reservoir simulator. - A domain-adapted language model for oil & gas terminology. - Award-winning computer tools to support qualitative decisions in geosciences. - A new generation of high-capacity interconnect switches for HPC. - Use of a professionally-run mentoring program to link junior researchers with senior managers. - Adoption of international best practices for managing partner involvement and translation between researchers and partners. - Establishment of robust collaborations with Germany, Brazil and UK. Potential Impacts: - Creation of a world-class centre of excellence in industrial information modelling and applications, linking the computer science and engineering disciplines across the process, energy and manufacturing industries. - Recognition of University of Oslo as a valued participant in European projects related to the digital transformation of industry.

Over recent years the sheer quantity of available data has been increasing at an exponential rate. The key bottleneck limiting end-user exploitation of this data is the difficulty of accessing just the relevant data. As one of the most information-intensive business sectors, the Oil & Gas industry is affected more than most; for example, Statoil estimates that its annual cost of data access is 250 MNOK in the Exploration unit alone. Recent technological developments promise a breakthrough in addressing the data access problem. The SIRIUS consortium assembles a team of research and development groups, from both academia and industry, with world leading expertise and a broad coverage of relevant technologies, together with two major Oil & Gas companies, and leading IT vendors. The consortium will collaborate in the development, deployment and evaluation of innovative data access technologies, following a value-adding innovation cycle designed to promote the transfer of knowledge and expertise as well as technology, and hence to accelerate the development and adoption of innovative data access technology in the Oil & Gas industry. Although SIRIUS is focussed on the Oil & Gas industry, the complexity of the challenges found in this domain mean that solutions developed here will almost certainly be relevant in other areas, and the involvement of major IT companies in the consortium will ensure the rapid transfer of SIRIUS innovations to other business sectors.