As powerful as digital technology is, it is only as good as the data—and mission—from which it draws.
The industry recognizes the potential digital transformation can bring to operations, but it is also aware that quality data is not enough. Speaking during the “Data-Driven Transformation: Leveraging AI and Digitalization for Operational Efficiency” panel at the International Petroleum Technology Conference in Kuala Lumpur on 19 February, industry experts talked up the possibilities that digital tech offers while urging companies to make sure they don’t fall in love with tech for technology’s sake.
Swarup Joshi, Microsoft’s director for energy and resources in Asia, said the best outcomes from deploying digital and artificial intelligence (AI) technologies require specificity about the problem that needs to be solved followed by a laser focus on adoption and embedding the tech.
“AI is a tool. There’s been tools that emerged in the past, and they come and go. Sometimes they get embedded. Ultimately, we have to use AI to deliver outcomes for the business, right, whether it’s for assets, for people, for customers,” he said.
He compared AI to an iceberg. With AI, he said, “what we can see is a lot of possibility, a lot of potential. We think about accelerating innovation or process optimization,” along with eliminating drudgery and increasing staff productivity, he said. “But, at the same time, as we go forward as an industry together, we’ve got to manage for a bunch of other things” that are below the surface, such as AI governance, data foundations, security, adoption and change management, employee skilling, responsible AI, moving from proving to scaling, defining success measures, and value realization.
Generative AI as a technology has already had a tremendous effect at the personal level, he said, adding that it can make processes and workflows faster or better. And looking to the future, Joshi said the “next wave of potential” might arise from bringing together different subdomains or data sets.
Krish Krishnan, general manager for KBB operations for ConocoPhillips Malaysia, said digitalization and AI have already contributed to business efficiencies the company has realized. Getting there, he said, started with a mission and required quality data.
“The opportunity is primarily on how we exploit digital technologies and that make our assets and processes more intelligent and automated,” he said. The differentiator is “how do we make our assets more automated, and how do we make it smarter to reduce the burden” on staff, he added.
High-quality and accessible data is critical, he said.
“Data itself is a whole topic, and it’s very important because it all starts with data. And then you go into infrastructure and platforms because you have to have the right infrastructure and platforms to deal with data,” Krishnan said. “Then, ultimately, once you sort that out … do you have the right algorithms to assess the data, analyze the data, and then make the right decisions?”
Along the digitalization path, it’s important to realize the tech is “a work in progress and requires further evolution and development,” he said.
David Reed, Weatherford’s executive vice president and chief commercial officer, said AI can generate value through the context of collect, contextualize, calculate, and control. Over the years, the industry has collected increasing amounts of data, which can be contextualized through unified data models. The data feeds into physics-based and empirical analyses calculations, and supervisory control and data acquisition systems control operations.
“The challenge we’ve had over the last 10 plus years is: Great, we have the data. But what do we do with it?” he said. “Data is only good if we understand what it is, and then we can make the decisions from it.”
In short, he said, it must be contextualized and used in calculations that ultimately improve operations. In one example that Reed cited from Australia, AI was able to help optimize production by 86% over 5,000 wells with negligible capital and operational expenditures. Before the AI-based workflow was in place, humans evaluated and made operations decisions on five to 10 wells a day, he said. The AI workflow can evaluate each of the 5,000 wells hourly and generate diagnostics and insights, he said.
“There’s a lot of opportunity to use data exceptionally well,” he said.
Porpich Pongparnish, vice president for digital at Chevron Thailand Exploration and Production, said one of the keys to working with technology, whether it’s AI, drones, or robot dogs, is to keep a focus on what matters.
“We need to start with the business value—the business value. Don’t fall in love with technology,” she said.
The operator has used AI in “nearly every aspect of the value chain,” she said and noted it’s helping understand the behavior of critical offshore assets.
AI and generative AI are evolving and how people work with these technologies must evolve as well, she said. One of the future opportunities she sees is connecting data and AI capabilities across multiple workflows.
“Right now, sometimes we just focus on a single workflow; but, there is a gap. There is a hidden secret between two or three workflows. How can we connect with the data and AI?” she asked.
The operator also plans to use other technologies in the field. Autonomous drone and robot dog pilots are expected to “help us inspect and understand the conditions” of assets and fields in the near future, she said.
Sinthu Satawiriya, head of ventures for PTTEP spinoff AI and Robotics Ventures, said that the industry “wants to have physical AI” instead of just processes and analytics. Physical AI would operate physically in the real world, he said, but the interface for that is important.
The company’s MARS wellhead robot for offshore operations has an arm that can take corrective action such as turning valves, he said.
The idea behind the robot, he said, is to “reduce the spread offshore,” saving money and reducing risk.
