Jim Fowler, CIO at GE, is on a mission to help generate $1 billion in productivity over a five-year period. One of the ways that productivity goal is being met is with an initiative called the digital thread.
The digital thread is a framework that connects data and processes from multiple systems together, making them easy to access and to integrate. The concept is not unique to GE, but the technology that enables the digital thread initiative for the company is. Most notably, GE's digital thread relies on data models called digital twins and the company's cloud-based industrial internet-of-things platform, Predix.
SearchCIO caught up with Fowler at the recent MIT Sloan CIO Symposium in Cambridge, Mass., where he talked about GE's digital thread and how technology is the backbone of the initiative's success. Below are excerpts from the interview; click on the player to hear Fowler's complete responses.
What's been an exciting undertaking and a big challenge for you in the last year?
Jim Fowler: The biggest challenge for me in the last year is taking on an initiative we call the digital thread. It's about linking processes in a different way and, specifically, the data and systems around those processes to drive productivity. We've taken on a $1 billion productivity target for the company over a five-year period. We delivered over $730 million of productivity in the last year through this initiative.
And it's about looking for the intersection points of processes. If you think about a thread of data that flows through your systems, it's being able to do things like connect an engineer who is designing a part to the services data related to that part, so they can make better decisions on design based on how something really gets used. We think this is going to be a fundamental change for the company, and we think it's where the productivity for the company for the next 10 years is going to come from.
How is technology enabling the digital thread initiative?
Fowler: There are two big technologies enabling us to go after the thread. The first is our platform, Predix, which allows us to connect machine data that is coming off of an aircraft engine, or a locomotive, or an MRI unit, with our enterprise data. It's an underlying framework; it allows multiple types of data and gives us a workbench on which we can build analytics and models.
The second big change is the adoption of what we call the digital twin. A digital twin is basically a data model -- it's a birth-to-death data model. And so for every asset that we sell a customer, we've started creating an electronic representation of that asset from the time the first part was ordered through its design, manufacturing and service. And that end-to-end electronic record allows us to then apply the models we've developed to help drive outcomes for our customers.
How did adoption of the digital twin emerge as part of this digital thread initiative?
Fowler: In our aircraft engines business, in our power business, we've been capturing machine data coming off of our assets for almost 15 years now. And we were using it for basic prediction of a failure. So, we were able to tell an airline when an engine part would fail, or we were able to tell a power company when we expected something within the plant or within a turbine would fail.
That was great, but what customers were asking us for is how we could apply that [data analysis] to the broader plan: Rather than just predict failure, how can it help me predict performance, predict an outcome or figure out how to get more revenue out of this asset?
It was that need of the customer that drove us to think about the fact that we had all this data -- not just the machine data, but maintenance records, consumer records, quality records related to these assets -- and that if you put them together, they would allow us to answer some of the [questions] our customers were asking.