the New Wave
exploration and development
achines capable of learning are the forerunners of the Fourth Industrial Revolution, as far beyond the vacuum-tube computers that heralded the Third Industrial Revolution as those contraptions were to the telegraph that wired together the Second.
The changes that have already come are revolutionary; the changes on the horizon are even more so, says Helena Wisniewski, UAA professor of entrepreneurship and chair of its management, marketing, logistics, and business analytics department. She’s also the university’s first Marion Porter chair, an endowed chair within the College of Business and Public Policy.
Wisniewski spoke recently at the Alaska Oil and Gas Association Conference on the topic of AI in the oil and gas industry.
“We are currently in the era of Generative AI. Generative AI is a subset of Narrow AI, which is specific tasks. Generative can create and design, and it uses natural language processing and deep neural networks to do that,” Wisniewski told the audience. “But we’re moving toward general AI. That’s where all the science fiction data comes in. General AI will have cognitive ability, understanding, decision making, and maybe reasoning—but we’re not there yet. Experts debate whether it will be in three to five years.”
What AI is currently capable of—and how it will change in the coming years—is something every industry should be looking at, Wisniewski says. Oil and gas companies are already tapping into some of the potential AI brings with it, and the industry is poised to expand its use even more.
If, for example, a widget factory creates a digital twin of the equipment, it would be possible to use a digital twin system to test (without much capital investment) how an upgraded logic control system might integrate with existing equipment and affect performance of existing machines, as well as the quality and amount of widgets that would be produced as a result of the upgrade.
At least one digital twin system is already in use in Alaska. Clay Koplin, CEO of Cordova Electric Cooperative (CEC), says his utility has an “extremely sophisticated” control system built up over the past twenty-five years. The cooperative hosted, in partnership with the US Department of Energy, the largest grid modernization project in the United States, Koplin says. As part of that modernization plan, CEC built a digital twin of its grid. It’s hosted at the National Renewable Energy Laboratory Advanced Research on Integrated Energy Systems campus in Golden, Colorado.
When a federal client asked CEC if it would be possible to extend power to a location 27 miles from the existing CEC power grid, an independent engineering firm used the digital twin to model and analyze whether the proposal would be economically feasible. The contractor was able to adjust parameters, tweaking it this way and that, and finally found a way to make the project feasible, Koplin says. Now the client is analyzing whether to move forward.
“We call it a one-stop shop, where you can get all the information about ACE and its insides that anyone could need,” says Yekaterina Novruzlu, senior instrument and controls engineer for the ACE project in an article on bp’s website. “Let’s say there was a maintenance issue with a part that is not readily accessible. In the past, it might have meant sending someone offshore, setting up scaffolding, and taking photos as a first step. The digital twin allows us to understand what is involved and then decide on a course of action in just a few minutes or hours.”
“We also looked at how, while maintaining and prioritizing safety, we could build in more time between maintenance shutdowns, known as turnarounds, when production has to be reduced or stopped,” Imanov adds. “The intention with ACE was that it would act as a pilot design for the platform of the future. When we started talking about this kind of automatization back in 2018, it seemed like we had a near-impossible task ahead of us. This start-up is a real milestone after years of conceptualizing, engineering, planning, and execution.”
Digital twins are not artificial intelligence, per se, but they are a model that can, when combined with AI, provide continuous learning, predict performance, and elevate predictive maintenance to a real-time strategy to more accurately anticipate and prevent failures, resulting in greater efficiency, safety, and reduced costs. Digital twins are being used not just to fix issues but predict them before they even happen, Wisniewski says. She enrolled UAA to be a member of the Digital Twins Consortium, which aims to accelerate the market by fostering development, increasing adoption, and improving the interoperability of digital engineering projects propelled by digital twins.
“Digital technology is a key enabler to facilitate the way we are doing business. As an energy company, we have to adapt to remain at the forefront of this transformation. The three previous industrial revolutions have demonstrated that not only the most advanced industries were successful, but the ones that were able to partner, develop, and work together effectively were the ones that often stood out,” says Christophe Vaessen, Shell’s general manager, commercial for its European Union, Middle East and Africa region, in an interview on Shell’s website. “Today, by bringing this OAI platform to life, we are building an open environment that enables all parties to work together toward a common ambition. The OAI is an open platform where companies can plug in and commercialize their apps. This includes not only international oil companies but also different sectors, such as cement or mining companies, that are running large operations and looking for digital tools to help with predictive maintenance.”
This way, similar user types aren’t having to reinvent the wheel, so to speak. The goal, Vaessen says, is to reduce integration and operating costs for users of the platform and to boost digital transformation in the heavy industrial sectors.
“The amount of time we all spend building our own proprietary platforms is truly remarkable. Working towards an integrated system is absolutely the idea behind the OAI. Now, the big challenge is that it will only work if you get adoption. So you can’t do this alone. And that’s why the fair value exchange is so important because it can’t be about one person gaining competitive advantage over the others. It has to be an ecosystem play,” Jeavons says. “Tech companies have done a very good job of what is commonly referred to as ‘coopetition.’ In many areas they cooperate, and in many areas they compete, and they’re able to do so simultaneously. Energy companies should consider adopting that coopetition mindset: in certain areas, it’s in our interest to cooperate for the good of society, whereas in other areas, competition is needed to drive innovation forward.”
In late 2021, Shell announced three new applications on the OAI platform: a process optimizer that “marries state-of-the-art LNG [liquified natural gas] process engineering and technology with data analytics” to optimize production; a corrosion advanced-risk modeling and analytics application that predicts internal corrosion and erosion to better pinpoint maintenance activities; and an autonomous integrity recognition app that processes data in the cloud coming from inspections made by handheld devices, drones, and robots to help inspectors evaluate issues and identify items that might have been overlooked, improving maintenance planning.
The initiative welcomed a cohort of new partners in 2022 and, in the same year, announced it had scaled its predictive maintenance program, driven by AI, to include more than 10,000 pieces of equipment across its global asset base, one of the largest deployments in the energy industry. According to Shell, the underlying technical infrastructure monitoring those pieces of equipment take in 20 billion rows of data each week, from more than 3 million sensors. It also trains, tunes, and runs nearly 11,000 machine learning models in production and makes more than 15 million predictions each day.
“AI is a huge opportunity for us. We’re going through a digital transformation, firstly to make ourselves more effective and efficient, and secondly to make sure we thrive through the energy transition,” Jeavons said in the Bloomberg article. “There’s huge disruption in the energy market, and as we go through this and move toward cleaner energy solutions, Shell wants to lead the way, and AI plays a huge role in that.”
OAI applications are not limited to traditional energy producers, he notes in the interview on Shell’s website.
“I see these solutions as applicable to the new energies business as well because reliability, integrity, and optimization are as relevant to wind farms, solar farms, battery storage, and hydrogen facilities as they are to natural gas plants and refineries,” he says. “It is an extremely compelling proposition to create an integrated platform that can help to manage the operations for a variety of energy and industrial processing companies.”
ExxonMobil isn’t the only company using AI to assist in exploration. Wisniewski says that type of application is an ideal fit for AI.
“Neural networks excel at recognizing patterns,” Wisniewski notes. “They can analyze seismic data to identify potential reservoirs, speed up reservoir identification, and increase the accuracy of reservoir predictions. I think you’ll [also] see an increase in AI-powered robots used for autonomous drilling, inspection, and maintenance operations, especially in remote and hazardous environments.”
As networks continue to rely on computer-based solutions, especially those hosted on remote servers, AI plays a crucial role in enhancing security, Wisniewski says.
“AI can play a vital role in predicting cyberattacks by detecting anomalies. Suppose an anomaly occurs in your system—how do you know it’s real? If you built your model on the physics of the system, and what [the attacker is saying] doesn’t track with the physics of the system, then you know it’s a bogus issue. You’re using AI to check against itself,” she says.
Wisniewski held a series of eight AI webinars through the College of Business and Public Policy from 2021 to 2024. One was on the topic of cyber security. She is organizing another AI-focused webinar series that will be open to the community, both in person and online.
“It’s a really exciting time ahead,” Wisniewski says. “The future is unimaginable, and there are technological advances coming that we have not yet envisioned.”