A utility company remotely observes that a tree is growing into its overhead lines, a robot makes a repair in a dangerous area of a refinery, and an oil company is alerted to a maintenance issue about to occur at a remote shale well. These are the type of activities that once were done solely by humans but are increasingly done using technology and analysis in the cloud.
The energy industry has always relied on data to make decisions, but in the last several years the industry has embraced cloud computing, discarding some of its massive data centers and looking to Silicon Valley for support.
"It's a rapidly escalating evolution. We've been talking about digital oil fields for 20 years. We've been working on these things. The thing that's changed is the databases and the way we can process those databases with analytics," said Jeff Shellebarger, Chevron's president of North America Exploration and Production.
Machine learning and artificial intelligence are being applied to everything from maintenance to exploration, and at the annual IHS Markit CERAWeek energy conference this week in Houston, companies like Amazon, Microsoft, Alphabet, Salesforce and other technology giants had a much larger presence then they've ever had before.
Amazon Web Services CEO Andrew Jassy addressed the conference and said the cost savings is the "conversation starter" and the cost savings are compelling.
The World Economic Forum predicted in 2017 that the digitalization of the global energy industry would bring a $1.5 trillion benefit.
"The cloud so perfectly fits their boom-and-bust cycles," said Bill Vass, vice president of engineering at Amazon Web Services. "Where things go up and down so the ability to scale massively when you need to, and scale down significantly when you don't, fits very nicely with the cycles you see in the oil and gas industry."
Vass said the speed of scaling up, or adding computing power, has changed the speed of some key processes. For instance, a small company had taken 100 days on its own systems to do a reservoir simulation, to see which way oil would flow and the best way to drill. "Now they're doing it in four days, to simulate all the different possibilities of how they would get more yield and make more money," he said.
In the past 18 to 24 months, Amazon, the largest cloud company, has seen an influx of companies from the more "conservative" industries, like oil and gas, health care and finance, according to Jassy.
"You could see a day when oil companies don't have data centers anymore. That is really the direction things are going," Vass said in an interview with CNBC.
Machine learning has been around since the 1970s, but it is being used more widely because of the computing power available to companies. "For smaller oil companies, a company like Woodside Oil can have as much compute power overnight as Exxon uses and they don't have to invest to do it," he said.
Big companies like Shell can quickly scale up processing power if there's a significant change in the oil price.
"The price of oil goes up, so they want to explore faster. They want to do production faster. They want to use machine learning to increase their yields, to improve their maintenance, to reduce their cost. They want to do all those things, so they can they can scale dynamically to the cloud, and then if the price of oil goes down and they want to reduce their expense, they can just literally turn it off," Vass said.
Energy companies have been under particular pressure to contain costs, but the move into digital technologies also is a differentiator.
"The whole technology wave was driven by costs. Now it's being driven by value creation," said James Rosenfield, senior vice president of IHS Markit and co-founder of the annual energy conference. "Most executives in the industry have come to the realization that technology is the single driver of competitive advantage in the future."
Rosenfield said now it's being applied to things like enhancing supply chain, reducing CO2 and it's attracting more young people to the industry.
"We've been in the data world forever," said Chevron's Shellebarger. "You think about artificial intelligence. You think about digital processing, data science, anything associated with that. We're really complex in terms of how we're tackling it. There's no one company that's a solution, or provider for that, so we have to see a portfolio of companies to put the pieces together to actually enable the business the way we're trying to do that."
Shellebarger said maintenance is one area that has evolved with technology. "We have maintenance out there so in a big oil field maintaining the wells, maintaining the equipment, maintaining the reliability of the system is a big piece of it. That's a very labor intensive position with lots of people looking at lots of different things, putting schedules together. We have decision support centers now that are automating a lot of those tasks," he said.
Technology also can send intelligence on what's actually going on at the wells.
"Think about a place like Kern River [California]. There's thousands of wells out there, Every one of those wells is wired up with sensors. Those sensors are all going into what we call a central decisions support center. They monitor the wells. We an adjust the wells, the pump rate on the wells and optimize that," Shellebarger said, adding the same technology is used on its steam system. "We're moving from managing those things by exception to actually getting into the artificial intelligence world," he said. "We're using things out there as sensors that didn't exist even two or three years ago."
Vass said there's a big cost advantage with the cloud.
"Since we started Amazon Web Services in 2006, we lowered our prices 72 times. Most of our customers, when they properly implement on the cloud, achieve between 22 and 45 percent savings," he said.
If a company installed a data center, it would still own it during an oil price downturn. "All that cap ex, you can't throw it out the window and sell it overnight if the price of oil goes down," Vass said.
Chevron sold its San Antonio, Texas data center to Microsoft last October.
Vass said the cloud is also enabling the energy industry to do more real-time streaming of applications, a growing area, and robotics. "We have a lot of oil companies that are working on removing people from the dull, dirty and dangerous jobs and optimizing automation at the refineries and on the rigs," said Vass.