The oil and gas industry has the data and the drive to pursue digital technology, but lacks a coherent strategy to fully digitalize its operations. Karen Boman reports from Halliburton's Landmark Innovation Forum and Expo (LIFE) in Houston.
Large oil and gas companies generate 1.5 petabytes a day of data, and even mundane operations such as transporting hydrocarbons can generate a terabyte of data for every 15000km, Dr. Sean Gourley, CEO of artificial intelligence (AI) firm Stealth Machine Intelligence and member of Anadarko Petroleum’s board of directors, told attendees at the 2017 Landmark Innovation Forum and Expo on 22 August in Houston.
However, the oil and gas industry falls behind other industries in terms of digital maturity, Gourley said, citing a 2015 study by the Massachusetts Institute of Technology’s Sloan Management Review and Deloitte, which gave the oil and gas industry a grade of 4.6 out of 10. The oil and gas industry only uses about 1% of the data it creates, said Gourley, and 70% of the time spent on that data is not focused on analytics, but finding, extracting, cleaning and modifying data to prepare it for analysis.
Despite the lack of digital maturity, Gourley said that he is optimistic that the oil and gas sector can reach the Holy Grail of digital technology – having seismic data be predictive of hydrocarbons phasing – as Silicon Valley’s focus on developing AI technology will create a step change in capabilities that will impact the oil and gas sector and other industries.
“It’s fair to say science around AI is accelerating, and we’ve only scratched the surface in terms of commercial applications,” Gourley said.
Gourley joined Anadarko’s board of directors in September 2015. Five years ago, Anadarko Chairman, President and CEO Al Walker recognized that many of the top-performing companies outside of oil and gas added talent from Silicon Valley to their boards. Walker decided to do the same; Gourley’s experience with Big Data and AI made him a good fit.
Walker was not seeking technical expertise from Gourley, but advice on implementing cultural change at Anadarko. This change was needed to ensure that technology was being applied in new ways, and that engineers in the field approached their work differently. Walker believes other oil and gas companies will add tech industry officials to their boards for this reason. In July, ConocoPhillips made such a move, appointing Sharmila Mulligan as a new director to bring technology expertise to the company’s board.
Besides the appointment of Gourley, Anadarko also rolled out over the past year its own data analytics group as it starts down a path toward using AI for seismic interpretation. Oil and gas companies can use AI to reduce breakeven costs while boosting efficiency. The group was not integrated into any one business segment; the idea was that the group would be a separate research and development group, with champions in different departments.
Advances in deep learning – a branch of machine learning, based on algorithms that can derive meaningful patterns from large datasets, and intended to move machines towards AI – are allowing machines to perform on the same level as human workers, or even outperform them. Through exposure to thousands of photos, these machines can be trained to detect objects such as cancerous tumors. In the case of oncologists, the best human doctors identify tumors with 96% accuracy, while machines can now identify tumors with 92% accuracy. But humans and machines put together identify tumors with 99.5% accuracy, Gourley noted. The same trend is true in weather prediction and chess. Using machines to train workers also can enhance their performance.
Geoscientists asked Walker if they will be put out of business by new technology. But, Walker doesn’t believe that will be the case. Gourley agrees, noting that, while some processes may be automated, the interaction of humans with machines will still be needed to solve the most difficult problems.
Gourley and Walker discussed how the oil and gas industry could learn from Silicon Valley on how to partner with service providers to accelerate the use of technology. Gourley acknowledged that Silicon Valley companies don’t always play well with each other. But, Silicon Valley’s ecosystem that allows companies to leverage learnings from each other has driven innovation. The only way for innovative products to move at a fast pace is through partnerships; no one company can do everything themselves. Even failures present opportunities as workers from failed Silicon Valley companies are often quickly recruited by other companies, sharing their lessons learned at their new companies.
Silicon Valley tech companies also are good at making bets on technology by acquiring companies that are truly transformational, Gourley said.
To attract the talent they need, oil and gas companies need to create an environment where the latest technology has a place at the table, and where new graduates feel they can actually build things, Gourley explained.
Students coming out of colleges like Stanford worry about not being at the top of their game. “The worst thing you can do for new graduates is to tell them not to build anything, just earn their stripes for 20 years before they might be put in charge of something,” Gourley said. “They want a place where they believe they will be pushed to be better in three years than when they came in.”
Image from iStock.
Read more of OE's Digital Technology coverage:
Internet of oil things (OE: August 2017)
The Digital revolution is now (OE: June 2017)
Teaching machines to speak drilling (OE: May 2017)
2020: A Digital Odyssey? (OE: May 2017)
Neural networking by design (OE: March 2017)