The ‘Battle for the Industrial Edge’ Unfolds

At last week’s OTC 2018 conference, I stopped at the Phoenix Contact booth for an update on its offerings. It has introduced some interesting products around industrial networking, connectivity, and electrical protection. What caught my attention, however, was the PLCnext controller, which is its ‘open’ PLC product line for the industrial edge.

Data science tools at the industrial edge?

The attention grabber was that its implementation of IEC61511 allows the inclusion of languages not associated with machine or process automation. Instead, the user can implement logic from third-party tools like MATLAB, Eclipse, or Visual Studio; tools of data science.

This is significant because PLCs are where manufacturing assets are and the primary path for most industrial data. In 2016, ARC wrote about the importance of this type of technology. At the time, we noted that edge/fog computing would be key to preventing a data deluge through a cloud-based system, deliver feedback locally and in near real-time, and address the concerns of customers who are reticent about sending data to cloud-based platforms.  Most edge platforms are focused on the network edge; smart routers and switches. Certainly, there are plenty of other products (like industrial PCs) that could compete or coexist within a cloud/fog scheme, but few are as intimate with sensors as PLCs and DCSs. We are now seeing a battle shaping up between classic IT companies and OT companies.  The good news is there is room for both.

Young Engineers and IT Will Lead the Charge

The push for this type of functionality at the control level will come from young engineers and the IT side of the business. Data analysts are unlikely to use an industrial controller for their work at the edge because of the lack of a modern installed base.  They are also less likely to know about the offerings that do exist. Industrial controllers, especially process controllers, have relied on less powerful processors (fewer transistors means fewer failures) to increase availability.  The available computing power is totally dedicated to control of the manufacturing process. This may well hinder the implementation of sophisticated controller-resident analytics in process control, or perhaps initiatives like Open Process Automation will prove that the process markets want the benefit of more powerful, flexible, and modern processors in the field to leverage analytic opportunities.   Discrete manufactures are likely to have a wider set of devices from which to choose and are more likely to implement.

End users need edge-to-cloud integration strategy

It is quite apparent that technology allows far more flexibility in architecture than ever before available.  Computing power is generally available in excess at all layers of the manufacturing process.  ARC recommends that end users don’t hesitate to pursue IT/OT/ET collaboration (and has done so for almost a decade). This facilitates the harmonizing of data access and visualization requirements as well as rationalization of where it’s best for the data to reside. This collaboration needs to come from the top and capital investment needs to be viewed more holistically.

Companies should already have a formalized vision for edge-to-cloud integration.  Due to the way technology is advancing, plans shouldn’t be restricted to today’s technology. Rather, the vision should be flexible enough to host functionality where it makes sense.  If the technology doesn’t currently exist to meet the vision, it will soon enough.