Back in 2013 Predix, GE’s operating platform for the Industrial Internet, was little more than an ambitious idea, albeit one that the company regarded as critical to its future. Predix launched that year in Chicago at the second Minds+Machines event, but, says Beena Ammanath, VP of data analytics for GE Digital, “we didn’t have many external customers attending … and four years ago Predix was more in PowerPoint than reality.”
By the time Ammanath wrapped her fourth Minds+Machines this November, everything had changed. There were “over 3,000 attendees and most of them were external to GE,” she says, and many of those customers and developers took to the stage to share how they’re working with Predix. “We can see the actual proof points of how companies are using Predix to drive their digital transformations, so that’s very exciting.”
GE has spent more than $US1 billion to develop Predix to become the backbone for the IIoT, the Industrial Internet of Things. By GE’s estimate, the convergence of machines, data and analytics could contribute $10 to $15 trillion in efficiency gains to the global GDP over the next two decades.
The power of collaboration via Predix to drive digital transformation for all sorts of companies was on show at Minds+Machines in San Francisco in November.
GE Reports caught up with Beena Ammanath—who is based in GE’s global software centre in San Ramon, California, is a member of Mensa and a board member of ChickTech—when she visited Sydney last month. “Predix is evolving into a data science platform,” she explains. “We have an Analytics Catalog with over 100 analytics … if you go to predix.io, there are 100 models that have been built by GE, within GE, for our industrial equipment, and our customers can buy it, use it as Lego blocks, to help them build digital apps.” They can even trial it for free for 60 days.
As Ammanath recently wrote on a blog, GE is seeing Predix out-of-the-box analytic microservices accelerate the app development for partners, citing one who reported app development three times faster than usual. Areas of recently added analytics in the catalogue include anomaly detection, data exploration and pre-processing, and machine learning, among many others.
Predix algorithms are supercharged because they have already been field tested gathering industrial performance information such as vibration-sensor data from gas turbines, wind-farm turbines, and jet engines.
GE has been using aviation as the prime example of Predix in action, but efficiencies are already being extracted in other industries. Ammanath points to ecosystem partner TCS, which is “using the analytic models on Predix to build out their apps … Predix gives them a platform where they cannot only develop their app, but deploy it to the edge, where the machines are. And we come in with all the domain knowledge from an industrial perspective, which they can leverage. They’re not starting from scratch.”
Digital Thread is another area in which Predix is getting to work. “We’ve started to take the data that’s captured in our manufacturing plants to predict when a machine might go down,” says Ammanath. “We’ve done it for two of our fully digital factories.” Now the plan is to use that digital factory as a template and make it available for use by any manufacturing plant in the world. Essential to GE’s vision for Predix is that it is equipment agnostic, and can be deployed on machinery and technology not built by GE and indeed, adds Ammanath, “beyond just the GE business verticals”.
At Minds+Machines in November, GE Digital’s CEO Bill Ruh took the audience for a tour around the company’s plans for a Predix-powered Industrial Internet.
There are now more than 19,000 developers around the world building apps on Predix, and field-services technology is another frontier in their sights. “We recently acquired ServiceMax,” says Ammanath, “and they are going to help us accelerate our digitisation process … we want to empower field-service engineers, whether they are maintaining GE-specific equipment or any equipment.” With an estimated 20 million field-service technicians at work around the world, developing digital tools that will give real-time information, such as what parts are needed and what work a machine requires, before the engineer arrives on site will make life a lot easier.
Ammanath says the overarching aim of Predix is “making the machines smarter to make human lives better. And as we are learning and evolving, with Predix we can help our customers take our lessons and then apply it.”
She gives the example of a small manufacturer producing school-bus transmissions, and nothing else. “It’s a relatively small company, and they don’t have the power of data science and big data like we do,” says Ammanath, who met with them at a Minds+Machines event. “They can use Predix and the apps that we’ve built to apply it to their manufacturing process for their engines, and help them drive productivity and digital transformation, without a whole data-science team or app-development team. That’s how it helps smaller companies: Predix works for any scale of industrial company and we have a whole ecosystem around it.”
Ammanath and her team have been immersed in the creation of Predix; now the potential is rapidly growing, in ways they didn’t even imagine. “What is exciting me is things that we are not thinking about, and they’re coming up as applications,” she says. “As I bring in the people from the outside world to open their eyes to all the possibilities, and what that reality looks like, it’s empowering.”