As the various applications of artificial intelligence (AI) gather pace, its potential to improve performance in the oil and gas industry is being explored. Two of BP’s technology experts explain
For many of us, the phrase ‘artificial intelligence’ brings to mind sci-fi films about robots that yearn to be real. From Spielberg’s 2001 classic A.I. to the more recent release, Ex Machina, these films are fantasy, futuristic and play a little to the populist fear that machines will one day take over the world. The truth is that, while AI is definitely science, it’s definitely not fiction: it’s real, and it’s already here. “It is taking over the agenda in many industries,” says Dan Walker, who leads the emerging technology team in BP’s Group Technology. “The automotive industry, for example, has witnessed AI companies take a leading role in shaping the future car, and many car manufacturers are now trying to catch up through major investments in AI. It is becoming obvious that the cars of the future will depend more on advances in AI than in classical engineering disciplines.”
Walker, along with colleague Paul Stone, a principal consultant in BP’s information technology and services, has been demystifying AI for BP. Stone explains: “The ultimate aim of AI is to develop systems that can take all those things that humans are good at - visual perception, understanding and communicating with natural language, adapting to changing situations, and complex decision making -and combine them with what computers today do best, such as large-scale, rapid number and data-crunching with consistency and accuracy. It’s about getting computers to learn and be able to reason without explicit programming mimicking the way the human works. At this point, we will have cognitive computing systems exhibiting human-like cognition and true thinking machines.”
This offers new opportunities for the oil and gas industry. “In the near future, there will be sensors with connectivity everywhere, recording data as often as you want them to and constantly creating new datasets,” says Walker.
“Using AI processes, known as algorithms, we’ll be able to combine datasets about areas such as flow rates and pressures and equipment vibration with data from the natural environment, such as seismic information and ocean wave height, to transform the way we run and optimize our operations.
Swapping nerve signals for algorithms
“The AI software architecture would be similar to our own central nervous system, which controls most of the things we do and the ‘hardware’ that we use in terms of arms, hands, fingers and legs, even though we don’t consciously think about it. Instead of nerve signals, AI uses algorithms.” Most areas of the oil and gas industry involve computer science problems that could lend themselves to AI solutions, Walker explains. “AI could help us to optimize well design and specify procedures to ensure that every well is drilled as efficiently and safely as possible, for example,” he says. “It could help us to improve equipment reliability and predict maintenance requirements of our facilities."
Fourth industrial revolution?
Stone adds: “We are already being proactive. BP’s Digital Innovation Organization is currently in the early stages of exploring the area of advanced AI and cognitive computing. We are looking at where AI and cognitive computing will have the most immediate impact and we are considering trials that will illustrate value.” The Castrol Brain is a pioneering digital AI tool that is initially ‘trained’ to answer marine customers’ common technical questions online through text message chat, but then teaches itself to respond better the more it is used. This will free up time for the Castrol team to focus on more technical and complex queries. “AI is enabling the fourth industrial revolution,” says Walker, “and it has the potential to help deliver the next level of performance.”