Last week at Hannover Messe, Microsoft and German carmaker BMW announced a partnership to build a hardware and software technology framework and reference architecture for the industrial internet of things (IoT), and foster a community to spread these smart-factory solutions across the automotive and manufacturing industries.
The stated goal of the Open Manufacturing Platform (OMP)? According to the press release, it’s “to drive open industrial IoT development and help grow a community to build future Industry 4.0 solutions.” To make that a reality, the companies said that by the end of 2019, they plan to attract four to six partners — including manufacturers and suppliers from both inside and outside the automotive industry — and to have rolled out at least 15 use cases operating in actual production environments.
Complex and proprietary is bad for IoT
It sounds like a great idea, right? As the companies rightly point out, many of today’s industrial IoT solutions rely on “complex, proprietary systems that create data silos and slow productivity.” Who wouldn’t want to “standardize data models that enable analytics and machine learning scenarios” and “accelerate future industrial IoT developments, shorten time to value, and drive production efficiencies while addressing common industrial challenges”?
But before you get too excited, let’s talk about a key word in the effort: open. As Scott Guthrie, executive vice president of Microsoft Cloud + AI Group, said in a statement, “Our commitment to building an open community will create new opportunities for collaboration across the entire manufacturing value chain.”
The Open Manufacturing Platform is open only to Microsoft Azure
However, that will happen as long as all that collaboration occurs in Microsoft Azure. I’m not saying Azure isn’t up to the task, but it’s hardly the only (or even the leading) cloud platform interested in the industrial IoT. Putting everything in Azure might be an issue to those potential OMP partners. It’s an “open” question as to how many companies already invested in Amazon Web Services (AWS) or the Google Cloud Platform (GCP) will be willing to make the switch or go multi-cloud just to take advantage of the OMP.
My guess is that Microsoft and BMW won’t have too much trouble meeting their initial goals for the OMP. It shouldn’t be that hard to get a handful of existing Azure customers to come up with 15 use cases leveraging advances in analytics, artificial intelligence (AI), and digital feedback loops. (As an example, the companies cited the autonomous transport systems in BMW’s factory in Regensburg, Germany, part of the more than 3,000 machines, robots and transport systems connected with the BMW Group’s IoT platform, which — naturally — is built on Microsoft Azure’s cloud.)
Will non-Azure users jump on board the OMP?
The question is whether tying all this to a single cloud provider will affect the effort to attract enough new companies — including companies not currently using Azure — to establish a truly viable open platform?
Perhaps Stacey Higginbotham at Stacy on IoT put it best:
“What they really launched is a reference design for manufacturers to work from.”
That’s not nothing, of course, but it’s a lot less ambitious than building a new industrial IoT platform. And it may not easily fulfill the vision of a community working together to create shared solutions that benefit everyone.