What's New in Production

Sam Charrington has some interesting things to say about artificial intelligence for industrial applications. The founder of CloudPulse Strategies said recently that artificial intelligence (AI), after an extended period out of the limelight, has returned to public consciousness in a big way. “These quiet periods…are collectively known in the industry as ‘AI winters,’ calling to mind the term nuclear winter, representing the quiet—and destruction in terms of research budgets—after the ash.”

“For better or worse, contemporary discussion of enterprise AI use cases has focused on applications in the digital domain. These are applications like getting people to click on ads, making recommendations, personalizing the customer experience, predicting customer churn, and detecting fraud of various sorts.

“But what about those parts of an organization whose operations extend beyond the digital domain? Surely they need AI too?” Here, “they” would be us—owners and operators of the hardware necessarily present at the “sharp end of the spear” in oil and gas operations.

What is industrial AI? Charrington defines industrial AI as “any application of AI relating to the physical operations or systems of an enterprise. Industrial AI is focused on helping an enterprise monitor, optimize or control the behavior of these operations and systems, to improve their efficiency and performance.”

A chain of related ideas orbits the broad concept of artificial intelligence. This includes machine learning (ML), about which Charrington says, “…while machine learning is only one way to build an artificially intelligent system, for all practical purposes, ML and AI are used interchangeably today.”

There’s also the more esoteric “cognitive computing,” which Charrington says is most often used interchangeably with AI. Also in the idea chain are big data, data used to train the machines, and predictive analytics. As applied to industrial AI, you can no doubt guess that the Internet of Things—more precisely, the Industrial Internet of Things (IIoT)—is at the end, literally and figuratively, of this chain.

The stakes are high. In a recent white paper, GE vividly distinguishes IIoT from the Internet of Things (IoT)touted by mass media: “The focus of the IIoT is not on connecting coffee pots to alarm clocks, but rather on connecting industrial assets, such as turbines, jet engines, and locomotives, to the cloud and to each other in meaningful ways.”

As Charrington points out, “Clearly the cost…is greater than the cost of Netflix showing the wrong movie recommendation, or Amazon upselling the wrong product. But the differences go further. This system is likely subject to any number of compliance requirements, and the system’s recommended action might trigger a variety of reporting actions. The development of the predictive model is likely significantly more involved than building a recommender: a variety of live and simulated engine sensor data must be captured; the sensor data likely require extensive cleaning before use; the model must be trained against the cleansed data; and it must be tested against a test data set, in simulation, and in production. This process likely relies heavily on a variety of subject matter experts, including systems engineers, maintenance and performance engineers, and more, not to mention the software engineering talent involved.”

Speaking of GE, the company has introduced a suite of digital tools it calls Predix, which it says is the foundation for navigating the “digital industrial transformation.” In the company’s view, this platform is where information technology (IT) and industrial operational technologies (OT) converge. In case you’re ready to transform, the company describes five steps: operating model and capabilities; data and connected infrastructure; partner ecosystem; culture change; and new business models.

Intelligent asset performance management. Assuming you’ve hit all the squares in this hopscotch game, the results appear promising. In one example germane to us, GE cites the success of an Australian operator needing to establish an intelligent asset performance management (APM) strategy that would minimize travel to well sites and allow most maintenance decisions to be made remotely from the company’s control center. Existing asset management strategy consisted largely of spreadsheets (with a variety of maintenance and operational data in silos); individual decision-making; actions with little corporate asset management strategy; and untraceable knowledge and activity.

The operator’s new APM strategy rolls up information to the company’s three primary focus areas—asset utilization, asset health, and asset stability—then parses it further for a management dashboard. The same information flows back down to various departments, ensuring that the correct engineers can be assigned and have the ability to drill down through the dashboards to find the location, the asset, the tag, and the issue to be resolved. Benefits to the operator are said to include minimized field travel; decreased operating expenditures; maximized productivity; and robust and sustainable operations.

A few more words about asset performance management, from GE: the company says it makes operations safer and more reliable while helping to ensure optimal performance at a lower sustainable cost. The firm goes on to say that APM enables intelligent asset strategies that balance three traditionally competing priorities—reducing cost, improving availability and reliability, and managing risk—to help optimize overall asset and operational performance.

The imperative to cut costs and improve profitability appears to have its most resounding response in “pressing the reset button” on the way things are traditionally done.