It has been over a year since Repsol and IBM announced their plans to jointly develop cognitive technologies that would allow human decision makers operating in the oil and gas industry to interact with computers more effectively, enhancing the industry’s ability to source, analyze, and filter big data to make it more consumable.
Building on what is commonly referred to as artificial intelligence, experts in this field are developing “super computers,” which are able to understand, learn from, and interact with humans. Many different companies in various sectors are currently developing cognitive technologies for specific activities, with the most pioneering being applied to the exploration and production (E&P) sector of oil and gas.
In an environment of low oil prices and ever tighter margins, cognitive technology can help energy companies increase the productivity of their oil and gas fields and minimize exploration risk when searching for new resources. The objectives of Repsol’s collaboration with IBM were twofold: to leverage cognitive computing capabilities to specifically help Repsol reduce the risk and uncertainty of future oil field acquisitions and to maximize the yield of existing oil fields—both of which can have a significant impact on the efficiency of global E&P operations of Repsol and other oil and gas companies.
Currently, offshore E&P is a capital-intensive exercise. The drilling of a single well can cost up to USD 400 million and the data used to determine the best location to search and drill for oil are often extremely limited and inaccurate. By developing these cognitive technologies, Repsol and IBM have been able to bring decision makers together, helping them to share insights, gather data sets from multiple sources more easily, and enable better, more prescriptive analysis. The ability to overcome limitations posed by big data has led to less uncertainty and, ultimately, reduced operational risk.
Optimizing Reservoir Production
Looking specifically to the optimization of reservoir production, a cognitive environment can adapt to the individual needs of a varied set of technical experts, equipping them with the tools needed to enhance their abilities to analyze data from varied sources. Technicians have also been able to tie in existing production models with the analyzed data and adjust them to more accurately match current production as time goes on.
Inspiration came in part from Repsol’s Excalibur project, launched with the objective of optimizing deposits. The project, tested in remote offshore fields in Ecuador and Brazil, was a combination of tools developed at the Repsol Technology Center. Excalibur uses mathematical techniques to comprehensively identify and evaluate deposits, optimize their development, and minimize risk. The tool enables a deposit portfolio to be quickly and accurately ranked and new investment opportunities to be identified that are difficult to detect using traditional techniques. In the tests carried out on a North Sea deposit, Excalibur achieved an improvement of 9% on the best solution published to date by other research companies and institutions.
Aiding Decision Making
The enhancement of the decision-making process for the acquisition of new oil fields, both onshore and offshore, entails strategic decisions across another group of technical experts. The purchase of new assets requires large financial investments based on very uncertain data. However, a cognitive system is able to interpret these data through the use of natural language processing. With the recent collaboration, a human expert is able to guide the system to quickly produce summaries for specific queries and run simulations that are able to inform the decision-making process and reduce the inherent uncertainty.
The technology being developed has drawn inspiration from another collaborative project, known as the Pegasus Project. The project has been able to improve the safety, efficiency, and profitability of upstream operations by constructing more fluid conceptual and geological models to explore hypothetical scenarios to guarantee the best decisions.
The objective of the Pegasus Project was to strengthen the capacities of technical equipment, through a much more efficient interaction between humans and computers. This tool is capable of analyzing hundreds of thousands of documents, reports, and data, to look for trends and connections between them, and propose the most appropriate hypotheses and solutions at any given moment.
The technology has undergone rigorous testing last year by a team of approximately 20 experts at IBM’s Cognitive Environments Laboratory in New York and Repsol’s Technology Center in Madrid. Thus far, Repsol has invested approximately USD 20 million to develop both applications, with early results expected during the first half of 2016. We hope that by the end of year 2, we will have a patented technology that can address the optimization of offshore and onshore fields, and that by the end of year 3, we will have a cognitive system that will assist in the acquisition of new oil and gas plays. We plan to launch several pilots during the course of 2016 and hope to apply the technology to all five continents in which we operate.
During the 3-year period, scientists will also experiment with a combination of traditional and new interfaces that are based upon gestures, robotics, and advanced visualization and navigation techniques. By using these techniques, researchers can leverage sophisticated models of human characteristics, preferences, and biases that may be present in the decision-making process. The technology will also introduce new real-time factors that should be considered when making decisions, such as current news events around economic instability, political unrest, and natural disasters.
Cognitive environments can look and feel very different—from boardrooms in the workplace, to cars, to homes, to mobile devices—but by being connected to one another, they can facilitate human-computer collaboration at “the speed of thought,” leading to more informed and robust decision-making processes. This new application will not only improve the way oil companies visualize and map oil plays, it will help them to make strategic decisions about the long-term development of their E&P activities. As we continue to develop the use of these cognitive technologies, we hope to be able to apply them to other areas of our business such as our downstream operations and refineries in order to help secure future energy supplies.
