Birthed from the marriage of physical and digital industrial concepts, Digital Twin is GE’s foundational analytic that aims to bring increased insight, understanding, and added value to GE’s big assets. Matt Nielsen, a principal scientist for GE Global Research, is leading a cross disciplinary team developing the Digital Twin, which much like any baby has required love, patience, hard work, and a little lost sleep.
Matt is a physicist and has been with Global Research for the last 16 years. He joined the Digital Twin project in December 2014 after the idea was gaining steam and looking to kick-off in 2015. Global Research senior leadership recognized the growing pattern of the physical world meeting the digital realm and realized its tremendous potential impact to GE.
The first application grew out of ongoing work with GE Aviation in life modeling of critical engine parts. The Digital Twin team was asked to use their technology to build tailored analytics for individual engines.
“With a successful reputation in both physical and digital asset develop, GE jumped at the opportunity to showcase innovation by blending the two worlds,” Matt said.
Digital Twin combines deep knowledge of each product’s individual parts and each part’s unique lifecycles with digital tools like data lakes, model infrastructure, visualization, estimation, controls, and specialized analytics such as operations optimization and maintenance workscoping. The outcome is a machine model that operates and updates itself from sensor data, providing asset-specific knowledge to help customers achieve better performance with lower maintenance costs. Rather than basing maintenance decisions on generalizations about a piece of equipment’s lifecycle, a Digital Twin will allow the asset operator to predict precisely when maintenance will be required based on the unique conditions experienced by that particular asset.
What GE is doing with its machines is not unlike the way online retailers create digital profiles of their customers and use data gathered from their online shopping habits to develop the ability to predict what and when they will buy. Similarly, GE is creating digital profiles of each machine – and each part that comprises it – in order to make smarter decisions about maintenance, optimize performance, and reduce unplanned downtime. Incorporating Digital Twin beyond GE’s Aviation engines is an obvious progression, as many GE machines like gas turbines, steam turbines, locomotives, and CT scan machines are comprised of individual parts that all need regular service and/or replacement.
“We’re seeing increased interest from GE customers as well as general industry trends, so we adopted an aggressive timeline to widespread deployment across GE’s large assets,” said Matt. “To achieve this we have many teams working together, united by our common focus to better manage risk and help customers eliminate unplanned downtime.”
Teamwork, Matt says, is crucial and is a big reason why the project is moving along so quickly. Another essential piece of the puzzle is an engaged and supportive leadership team, which helps bring vision and guidance to the program pathway. Matt’s role is working day-in and day-out to coordinate meetings, motivate progress, and navigate barriers to ensure a collective vision. Plus, he has a lot of fun getting involved with the cool technology being developed by the team.
“It’s a matter of bringing together people from different disciplines and allowing them to do what they do best — create,” said Matt. “Then we can take the individual pieces and do what GE does best — innovate.”
In the end, digital + physical will equate to a whole new level of performance for GE’s assets, an outcome that all of GE is anxious to see.
“The program is generating a lot of excitement and interest is continually growing. Digital Twin is evolving as we rise to meet the technical challenges posed by both our internal and external stakeholders,” said Matt.
Work on GE’s Digital Twin is ongoing. In fact, there’s probably somebody, somewhere, at this exact moment pouring their heart out into physics-based modeling. The goal is to rapidly deploy Digital Twin to many of GE’s large assets in 2017. Stay tuned…