Industrial companies are embracing digital twin technologies, with a view to keeping costly assets in good repair and maximising their useful lives.
Out on the waters of the Gulf of Mexico, some 25 miles south-west of Port Fourchon, Louisiana, the Noble Globetrotter I drilling vessel is hard at work. The ship is drilling for oil under a 10-year contract between its owner/operator, offshore drilling contractor Noble Corporation, and oil giant Shell.
Meanwhile, over 1,000 miles away, in Lisle, Illinois, a digital representation of the Noble Globetrotter I is being closely monitored by engineers working at General Electric’s Industrial Performance and Reliability Centre (IPRC). This is the vessel’s digital twin.
The IPRC runs on a 24-hour basis, seven days a week, with the goal of spotting problems and increasing uptime for GE customers’ industrial assets, in fields as diverse as oil and gas, power generation, mining, and aviation.
The Globetrotter I’s digital twin is based on data collected from physical assets found on the ship: specifically, its drilling control network, along with its power management and dynamic positioning systems. Each of these assets has its own sensors and control systems, which provide steady streams of data. These are harmonised and centralised on the ship, before being transmitted in real time to the IPRC.
There, the Digital Rig solution developed by GE and Noble gets to work, applying advanced analytics to data models based on the digital twin, in order to detect unusual behaviour patterns, which might indicate a problem with equipment onboard the ship.
The digital twin concept is not a new idea, and could be seen as a useful byproduct of CAD. However, the use of live sensor data to model real-world objects seems to date as far back as 2002, to a presentation to industry given at the University of Michigan by Dr Michael Grieves. In his speech, he suggested that a digital representation of a physical system could be used to monitor and support the entire lifecycle of its physical sibling, in order to keep it functioning.
But in recent years, as sensors have become cheaper and the cloud has delivered almost limitless, low-cost storage space and processing power, the idea has really taken off.
Digital twins were one of Gartner’s Top 10 strategic technology trends for 2018, with analysts at the firm predicting that organisations will implement digital twins “simply at first, then evolve them over time, improving their ability to collect and visualise the right data, apply the right analytics and rules, and respond effectively to business objectives.”
Already in 2018, Internet of Business has reported on plans at IBM to create a digital twin of the Port of Rotterdam, Europe’s largest shipping hub.
Then there was the news last week that professional car racing squad Team Penske is teaming with Siemens to create digital twins of its vehicles, enabling engineers to simulate engine configurations, develop new parts, optimise performance, and even predict race results.
And Kärcher, manufacturer of vacuum cleaners and pressure washers, has said that it will use software from Dassault Systemes to create digital twins for “system engineering, configuration, manufacturing, after-sales services and packaging design.”
But back to the Noble Globetrotter I. According to GE, Digital Rig has already “captured multiple anomalies” in its physical assets and produced alerts about potential failures up to two months ahead of when they might otherwise have been expected to occur. The goal is for the solution to deliver a 20 percent reduction in operational expenditure across targeted equipment, through high-octane predictive maintenance.
These kinds of results can only come from digitalisation, claims Krishna Uppuluri, vice president of digital product at GE Digital. “If you look at the way that drilling contractors operate, it’s been much the same style for the past thirty to forty years,” he says. “It is predominantly based on experience, gut feel, and calendar-based maintenance.”
In other words, problems are diagnosed on the basis of hunches, and maintenance schedules are strictly observed. Some equipment undergoes maintenance even when it’s running fine, just because it’s due to be inspected.
But now when something starts to go wrong, says Uppuluri, the team at the IPRC can spot it early, and package up an alert with relevant data to send to the vessel’s crew, out on the Gulf of Mexico. This way, they get plenty of warning if they need to order and transport new parts (and experienced engineers to fit those parts) out to the ship.
GE and Noble plan to build digital twins of four different drilling vessels at first, says Uppuluri, and then use these as the basis to roll out the technology to the rest of Noble’s 28-strong fleet.
This kind of project may be just the start. Says Gartner analyst David Cearley: “Over time, digital representations of virtually every aspect of our world will be connected dynamically with their real-world counterparts, and with one another, and infused with AI-based capabilities to enable advanced simulation, operation, and analysis.”
What we’re seeing, he reckons, is a long-term shift to a “digital twin world” – with huge implications for all kinds of professionals, from city planners and digital marketers, to healthcare workers and industrial planners.
Internet of Business says
Perhaps the biggest digital twin programme currently in existence can be found at CERN in Geneva, where the 27km loop of the Large Hadron Collider remains the largest machine ever built. Every component in the LHC – and on the CERN campus, which is the size of a small town – is logged in an enterprise asset management (EAM) system as a digital twin. This enables engineers to keep the big science running, and for repairs, upgrades, and replacements to be planned for well in advance.
And the system has another, equally important benefit: in a 27km complex full of expensive equipment, the digital twin system also tells engineers exactly where the tiny bolt that needs replacing is located. That’s not to be sniffed at when a round trip on a slow maintenance vehicle may take several hours.