While many of my colleagues have recently written blog posts to assess their 2016 predictions and forecast what key trends will emerge in 2017, I made no 2016 predictions to review. I spent most of 2016 looking at everything from ERP to the moving Cloud to impact-technologies, such as Industrial Internet of Things (IIoT) and Big Data. So, instead of making predictions about what could or will happen, I’ll use this opportunity to say what I believe manufacturers and other end users should do in 2017.
If there is one professional resolution your organization needs to have in 2017, it is: "In 2017 we will create (or update if it already exists) an Operational Architecture plan that recognizes that plant floor (or operational technology) is as much an intrinsic part of our business as utilities, raw material, and labor."
Operational Architecture Wasn’t as Critical in 2016
At the end of 2015, looking forward to this year, LNS was talking about Operational Architecture, explaining what it was, how it related to Digital Transformation, and how it helped with the issues related to deploying IT and OT in operations. We took the position that forward-thinking companies looking to pursue Digital Transformation SHOULD include Operational Architecture as part of a framework to accelerate their results.
Part of the reason we were not insistent on companies putting Operational Architecture planning at the top of their 2016 to-do list was that at the end of 2015 there was more hype than reality about Big Data and Predictive Analytics, the IIoT, and Cloud technology on the plant floor. At roughly ½ the user group events we went to in 2015, IIoT was either not a key topic or if it was mentioned it was in the context of remote monitoring, much as industry had been doing previously. Some vendors were even painting IIoT as mostly hype. In those vendor events where IIoT was a focal point, it was more about strategic vision and direction than hard value proof points.
What’s Different in 2016
In 2016 there was a decided shift in the dialog. Every single automation vendor used 2016 events to put forth their IIoT vision with some highlighting successful pilot projects for full-blown IIoT platforms. Also in 2016, we saw Cloud clearly overtake on-premise for new ERP installs in several vendor communities as well as the introduction of Cloud-only MES solutions to the market. Finally, more Predictive Analytics vendors emerged from “stealth mode” than in any of the previous years. In addition, there were numerous third party and media sponsored IIoT events, more than ever before. With so many alternatives now in the market and so many varying approaches by the vendors, users face some critical decisions.
Why Operational Architecture Planning is a Must-Do in 2017
A flurry of announcements in the fall of 2016 at vendor conferences revealed some vastly different approaches to IIoT platforms, Big Data & associated Predictive Analytics, and Cloud. Many automation companies, which have been leveraging Microsoft technology for their HMI, Programming, MES environments, and other non-real-time applications are lining up behind Azure as their Cloud delivery platform, with Azure and Cortana analytics as the engines behind their predictive asset oriented solutions.
Others have struck out on their own path with a higher-level platform that is inclusive of Azure but not dependent on it while others have lined up behind a multi-platform approach that supports SAP, IBM, Azure, or other solutions. On the Business Applications (like ERP) front there is also multiple approaches. Some are going full-on Azure, Others like SAP and Oracle are going their own way while others like Infor have mate Amazon’s AWS their platform.
The implications for end users are serious. Select the wrong combination of automation, MES, and ERP vendors, and then decide to go with a Cloud-based strategy and you may find yourself having to become adept at existing in Azure, AWS and an ERP vendor’s proprietary environment. Adopt a multi-vendor automation approach without considering their IIoT platforms and find yourself having siloed systems that cannot coordinate, potentially devastating if they are sequential machine centers that need to be synchronized. Forget about implementing a RCM approach to APM if you can’t get at and use the information from those machines as well.
With IIoT standards still evolving, platforms still in definition making decisions about vendors and architectures can’t be left to chance. Manufacturers MUST have a plan. They need to understand where they are today from a technology perspective and where they will need to be in the future to remain competitive. Without an Operational Architecture, there is the probability that over the next two to three years they will spend heavily on technology that they thought would enable them to transform into a Smart Manufacturing enterprise. Instead, it will create a tangled web of technology that puts them at a competitive disadvantage.
While a few may succeed at applying “incremental decision making” (better known as muddling through), the majority will find themselves with dysfunctional systems that put them so far behind their competition they must start all over at an expense that may well doom their business. The way to avoid making these mistakes is to embark on an Operational Architecture planning effort in 2017, or if you have done one in the past, refresh it in light of the technological changes in the market today.