Technology’s disruption in the world of oil and gas has been most pronounced in recent years as companies learn how to operate with fewer people, though production still surges.
Big Data and analytics – more importantly – how companies are using their Big Data and analytics is integral to their future success.
“There’s really two types of data – data that informs a business problem and data that does not,” Michael O’Connell, chief analytics officer for TIBCO, told Rigzone. “You can collect all the data under the sun, but if it’s not relevant – no matter how much of it you get – it’s not useful.”
Technologies such as Artificial Intelligence (AI), machine learning and Big Data are already disrupting the oilfield and could be instrumental in hot spots like the Permian Basin.
“We’re going to put 110,000 wells in [the Permian] in the next couple of years. If you add up all the man years we have in the major operators that are going to be doing that, there’s just not enough man years to do it,” Troy Ruths, CEO and chief data scientist for Ruths.ai, told Rigzone. “So that gap is filled with technology, and particularly for unconventionals, it’s data. Because the reservoirs are so complex, the processes need to move faster, and companies are realizing that data is a critical element.”
Still, it’s been estimated that oil and gas companies are using only 1% of their data.
“Picking up all this data and not using it is kind of like building a Facebook page and you’re the only one to look at it,” Robert Golightly, senior manager, product marketing, manufacturing at AspenTech, told Rigzone. “What’s going to make a difference is finding those behaviors – whether innocent, malicious or anywhere in between – that degrade equipment. If you can find a behavior, that’s where all this expertise really starts to become more important than it really is.”
Ruths said the big challenge he sees in oil and gas is people who are starting to operationalize their analytics.
“Now they have larger user bases and more workflows and more data that needs to connect to it,” he said. “You hit different growing pains when you try to operationalize something versus when it was in the back office untouched by large groups of people.”
TIBCO creates custom solutions for the energy sector that allows companies to extract data from difficult systems and integrate it for analytics.
O’Connell identified the following three market forces:
As the industry utilizes Big Data and analytics more heavily, the natural question to ask is what kind of human capital it takes to apply those technologies. A few years ago, the industry was looking to hire more data scientists to account for the large amounts of data companies were gathering.
Today, technology helps with that.
“Machine learning, without having to be a data scientist, lets engineers look at historical data and figure out what’s really going on,” said Golightly.
He gave the following example: if there are two pieces of equipment – one in reserve – and periodically one is shut down to bring the other up in its place, it’s in that transition where there is carryover behavior occurring.
“The software helped them figure that out, but it didn’t just tell them the answer. It gave some really good clues, so people with strong domain expertise could look at it and say, ‘maybe it’s this,’” Golightly said. “That curiosity led them to identify the process behavior that was causing the problem and actually change the operating protocol.”
Golightly went on to say the skillsets the industry requires can be a bit much: understanding the infrastructure, the systems, analysis, physics and chemistry of the process is a lot to ask.
He said machine learning can allow somebody with domain expertise to be relieved of that burden of having to know everything, instead giving them clues to figure it out.
“There’s always going to be that sort of bespoke analytics that is going to cry out for that data scientist and that specialized team of people, but that can’t be the factor for your analytics strategy because there’s only so many of those guys to go around and they still aren’t equipped with the domain expertise,” Golightly said. “Everybody is looking at the data scientist as a silver bullet. I don’t think it’s sustainable. It’s hard to compete against Amazon and Google [for that talent].”
There’s also a cultural change that needs to happen within organizations for these new technologies to be successful.
“Executives aren’t getting the traction they want from their investments because they weren’t the change agents of the culture that goes with it,” said Golightly. “If you’re going to create a truly cross-functional organization, you have to create a culture that makes it worth it for them to get out of their silos and comfort zones.
Golightly said it’s all about culture and technology is probably secondary.
“Until the industry packages the technologies in a way an everyday engineer can use, it hasn’t done itself a favor other than create scarcity by predicating its success model on rare skillsets,” he said. “It just seems kind of foolish.”
For Ruths, culture is for humans and it entails a lot of training.
“What do the humans need as we talk about all of this machine intelligence? What we think our users really want is training and content,” said Ruths.
The industry downturn, rise of unconventionals and the Great Crew Change means the cultural shift is ready to happen in the industry.
“The cool thing about all of this AI and Big Data is that it’s ready and it’s going to work,” Ruths said. “I think it’s going to be a really exciting next five years for oil and gas. I would suggest everyone become more adept and focused on their data because that is really where the opportunities and resources lie.”