Extracting Value from Data Using an Industrial Data Platform to Provide a Foundational Digital Twin

Abstract

Oil & Gas data currently exists within a world of data silos. Lack of data is not the challenge. A wide variety of data is collected, including sensor values, P&IDs, ERP, and depth-based trajectories. Rather, the challenge pertains to data usefulness. The root of the problem is a combination of factors, including poor data infrastructure, incompatible operational data systems, and restricted data access. All this translates to a low maturity of digitalization across the Oil & Gas industry. To date, digitalization efforts have been limited to pilot projects, proofs of concept and case studies, with no large-scale operationalized projects. Aker BP, one of Europe's largest independent Oil & Gas companies, has broken through the typical roadblocks by deploying an industrial data platform across all five of its operational assets. The platform aggregates and processes data from sensors and contextualizes it, structuring it in relation to process diagrams, production information, 3D-models, and event data (maintenance, incidents). Everything linked in the real world is also linked in the platform. This has dramatically reduced the cost of integration and maintenance, while simultaneously enabling scalability, speed of development, and data openness throughout the Aker BP organization. The data platform handles live and historical data for close to 200,000 sensors, with a peak transfer of 800,000 data points per second. Internal and external experts are able to apply state-of-the-art algorithms to visualize and solve critical business problems. A range of third-party applications and data scientists also use the 1+ trillion data points in the platform to create value and support Aker BP's strategy for day-to-day operations and long-term digital transformation. To realize the promise of digitalization, unlocking the value of data must be made a priority within the Oil & Gas industry. This paper will describe the implementation of the industrial data platform, explaining how data streamed from many, disparate, underlying systems is contextualized in the data platform to provide a holistic view of all processes and operations, thus creating a foundational digital twin for each asset, ready to empower machine learning applications for optimization and automatization, as well as human-facing applications, such as advanced visualizations and apps for the digital field worker.