People living around the port of Carrara, in Tuscany, Italy, are used to seeing giant slabs of the region’s signature white marble moved onto ships bound for every corner of the earth. But in the dead of night late in June, they witnessed an even bigger spectacle — a slow-moving, 3,500-ton turbogenerator built by Baker Hughes, a GE company (BHGE), headed for western Kazakhstan, where it will be one of five generators providing power to the Tengiz oil-extraction project.
At 60 meters long and 25 meters high, the huge machine required a custom-made “self-propelled modular transporter” to make its way from the assembly plant, through the town and to the port. But the effort was worth it, as having a fully built generator arrive (on its own boat) at the project will save Tengiz millions of dollars and countless hours of labor it would have had to spend assembling the machine locally.
It was almost the last phase of a logistical supply chain that stretched thousands of miles and many months from the time parts were ordered to build the generator to the final delivery. Orchestrating the many moving parts needed to build and deliver this kind of turbomachinery requires supply chain management precision. The order and delivery cycle for such complex projects can span years and requires bespoke components. So if a customized compressor or steel beam is delayed, it can bog down the entire project for weeks if not months. Meanwhile, other parts that are waiting to be used in the project would wind up sitting around, running up inventory costs. “You can’t go to Home Depot and pick up a part for these systems,” says Peter Koudal, supply chain technology leader at GE Global Research.
BHGE’s involvement in oil and gas projects ranges from the manufacturing of compressor trains that help turn natural gas into liquid — they can weigh as much as four double-decker passenger jets — to subsea machines for projects located in remote areas of the ocean. To solve the supply chain problems associated with these projects, GE Global Research and BHGE formed an interdisciplinary team that enlisted digital twin technology to help. “The delivery and construction of these modules is incredibly complex,” explains Davide Iannucci, vice president for global projects at BHGE’s Turbomachinery & Process Solutions (TPS) business. “Just imagine the precision needed to place the upper section of the main gas turbine generator module – weighing around 1,000 tons – on the lower one, with a maximum margin of error equal to the thickness of a few sheets of paper!”
Digital twins allow engineers to create a virtual model of a part, a machine or a process. GE already deploys more than a million digital twins to make other aspects of its business more productive, using them to build more efficient wind turbines and to determine when jet engines need servicing.
But by applying digital twin technology to the large, lengthy and complicated supply chain processes, GE and BHGE are breaking new ground. For the most part, supply chains have been deterministic in nature, meaning that outcomes were planned based on predetermined calculations that might be updated monthly or quarterly. But in reality, supply chains never operate as planned — too many factors, such as changes in lead times for supplies and part delays, impact how they function on a daily basis.
With digital twin technology, engineers can manage and track the supply chain from the beginning of a project to the end with a single analytical system that’s viewed in real time and continually updated.
Annarita Giani, a complex systems scientist with GE Global Research, says that before digital twin technology, it might have taken engineers weeks or months to manually gather and analyze information related to a supply chain problem. If a solution was found, it was often obsolete because it was based on already antiquated data. With digital twins, plans are continuously updated and instantly shared across multiple organizations, giving engineers and materials managers the ability to solve an issue as soon as it arises.
“We can see what needs to be seen and optimize the process from order to delivery,” Giani says. She would know. She worked closely with business leaders from BHGE, including packaging, assembly and materials managers Andrea Tirinnanzi, Giuseppe Severi and Antonio Cometti. “Without their commitment and expertise, the digital twin engineer-to-order project would simply not have been possible,” Giani adds.
When building the digital models, researchers collect a vast amount of information from hundreds or thousands of sources, including individual suppliers who might manually input information into the system, sensors embedded into specific parts and software codes that retrieve specific records from databases. Software then compiles the data into a single system that can be modeled and analyzed in real time to determine how the supply chain is currently performing and what the high-priority issues are. Such visibility allows supply chain planners and engineers to troubleshoot when snafus such as foul weather, stalled trucks or delayed parts disrupt the process. All of it allows GE and BHGE to improve delivery times, reduce inventory costs, monitor material shortages and create other efficiencies within the system.
Such optimization creates efficiencies in the flow of goods and the use of existing inventory. So far, the digital twin technology has helped BHGE’s Turbomachinery & Process Solutions unit cut its inventory costs by more than $30 million and increase inventory turns by more than 100 percent.
The digital twin is one of the reasons Frost & Sullivan’s Manufacturing Leadership Council recently gave GE and BHGE the manufacturing leadership award for supply chain innovation.
“You have to be able to look at what’s the best plan every day, not just every month or every quarter,” Koudal says. “That’s the goal and that’s what we’ve enabled.”