For several years, the oil and gas industry has experienced a drain of experienced personnel, creating a gap between very experienced people and lower-level resources.
"This demographic issue has affected the industry, since the knowledge retained by these more experienced individuals has historically been a blend of 'art and science,' and has been relied on to make decisions in the exploration, production, trading, and refining of petroleum and petrochemicals," said Tony Mataya, director of the Energy Practice at technology consulting firm Information Services Group (ISG).
One potential way to deal with the challenge is to deploy cognitive technologies. "Cognitive systems partially address this brain-drain problem by capturing the thought processes and decision-making capabilities of senior personnel that may soon exit the business," Mataya said.
There are other opportunities for oil and gas companies to leverage cognitive computing. For example, by combining information, including data from diverse sources, experience, and rules of thumb, teams at oil and gas companies can make decisions based on incomplete information and historical trends.
"Cognitive computing helps solve the problem by capturing vast amounts of data from varied sources, including the knowledge from experienced engineers and technical staff, to help oil and gas companies understand trends and previous results, match similar patterns, and connect disparate information to support decisions," Mataya said.
Like many other industries, oil and gas has been slow to adopt cognitive technologies such as artificial intelligence (AI) without a specific tangible application area, Mataya said.
"Cognitive is typically considered a subset of AI that deals with cognitive behaviors associated with thinking," Mataya says "Since many of the application areas in oil and gas require intuition combined with information to make incredibly risky decisions, applying AI and cognitive concepts increases the quality and accuracy of decision making." It does this by using more complete sets of data, considering more iterations, and connecting information previously thought to be unrelated.
Among the key challenges of adopting cognitive technologies are data quality and availability, disparate data types and formats, lack of skilled resources, and cultural factors.
"Data considered for many of the processes requires scrubbing to remove errors or unnecessary data, and comes in different formats," Mataya said. Companies need to understand the data required and get the information in a format for use in cognitive systems.
There's a need for skilled resources that understand both the data required and the cognitive technologies themselves. Matching up skilled technology resources with industry experts is essential to success.
"The number of industry experts have decreased while the demand for their knowledge-based experience is increasing," Mataya said. "The oil and gas industry has historically relied on a blend of thinking, intuition, and ever-increasing amounts of complex data to make decisions."
Data is available in more useful formats, with visual representations leveraging higher and higher computing resources. But there can be a lack of trust in having technology take over more of the thinking related to decision-making, rather than relying on human experience and intuition, Mataya said.