[sidebar id= 2]
The concept of “open systems” first became part of the process control vernacular around 1990, when every distributed control system (DCS) maker on the exhibit floor of the once-massive ISA annual symposium trumpeted its support of openness. Depicting competing “closed” systems as dinosaurs facing extinction or worse, the lustrous marketing banners gave way to the reality of what passed as open communications: support for Modbus RTU. That the extremely expensive and custom-manufactured systems of the day supported Modbus wasn't a bad thing, even though it was often a costly addition and complicated to configure.
In the decade that followed, the early efforts to create a standard for all-digital field communications produced ISA’S SP50 specification that later became the core of FOUNDATION fieldbus (FF) and, to some extent, Profibus PA.
The protocols accomplished a number of key objectives: all-digital communication and power over a single, twisted-pair cable (even the existing cable being used for 4-20 mA analog instruments), hazardous area capability and (with FF) prioritized messaging that provided time-sensitive, deterministic, publish-and-subscribe relationships, as well as asynchronous, lower-priority data exchange for diagnostics and the like.
The specifications were an added burden for device manufacturers—many of whom had limited interest in providing more than a measurement. Meanwhile, system suppliers had to devise engineering tools to configure and manage what amounted to a foreign control system. Despite tightening specifications, testing and certification, the need for a determined, in-house champion to anticipate and quickly resolve communication, configuration and control issues didn’t go away. Concurrently, end-user corporations were choosing to thin the ranks of their technical specialists.
Many end users found their experience with all-digital standards for device integration untenable, especially as support faded. For subsequent opportunities, justifying the effort and discipline necessary to make it a strategic choice for device integration was a struggle. Unlike the simple, ubiquitous and inelegant Modbus, FF and PA haven’t become de facto standards.
The Ether-bus asteroid
Ethernet is arguably—undeniably?—a de facto standard. And today, we have Advanced Physical Layer (APL), a single-pair Ethernet physical layer providing communications and power to field devices in hazardous areas. With present architectures, one needs controllers, gateways, HART multiplexers and like appliances to gather device data (including process variables), and pass them to other subscribers and clients. Imagine all those interposing devices being replaced with secure Ethernet switches.
Will one even require I/O cards for future device-to-controller communications? With latencies that are miniscule relative to process time constants, why make device manufacturers bother with more than rudimentary signal conditioning, i.e., convert a differential pressure to a flow and assign engineering units? Ethernet connecting field devices have the potential to simplify system engineering enormously, and conceivably use a completely vendor-agnostic standard.
We’ve heard that the system supplier community was moving away from manufacturing and selling hardware since the 1990s. But for years, the road forward has been paved with proprietary hardware and network offerings. One wonders whether purveyors of hardware ecosystems incompatible with such openness will see APL as friend or foe. Are the fortunes of system suppliers too tied to hardware? Why not just offer Profinet, EtherNet/IP or vendor-specific, proprietary Ethernet communications over APL?
On the other hand, if your DCS adds support for OPC UA and HART-IP, APL has the potential to wend its way into brownfield sites. A piecewise migration of increasingly critical applications is not only conceivable but likely. If barriers to digital integration of field devices dissolve, perhaps the age of dinosaurs will finally end.
[sidebar id= 1]
