- By Darrell Halterman
- Factory Automation
Well-designed edge controllers remain highly compatible with existing PLC/PAC ecosystems. Embrace new strategies and tools to adapt and gain new capabilities.
Edge computing, new communication methods, and advanced analytics are among the hottest topics related to industrial automation, a field more often associated with stoic reliability than cutting-edge technology. Yet there remains a fundamental need for creating traditional deterministic control solutions suitable for new and retrofit automation projects.
Designers are often concerned about managing interactions with industrial automation devices, especially as hardware and communications capabilities rapidly evolve. Fortunately, by adopting some basic tools and strategies for abstracting application data, interactions with modern edge controllers can be nearly as easy as traditional programmable logic controllers (PLCs).
As hardwired relays yielded to PLCs, which in turn were surpassed by programmable automation controllers (PACs), the means and methods of programming control systems have adapted. The same evolution is occurring as edge controllers and other edge devices begin entering service. Controls programming during the edge evolution does not have to mean wholesale change but can instead offer more choices.
The edge is here
Most users have heard terms like “edge computing” and “digital transformation,” but many may feel unsure how edge computing can solve problems in their applications. At least, users stand to gain improved visibility into equipment and process operations, even via mobile devices. Streamlined data gathering capabilities support analytics at the edge and in the cloud, which provide novel insights for improving availability, utilization, and efficiency. In the most advanced scenarios, edge controllers can autonomously leverage analytical results to advise deterministic control applications for near-real-time responsiveness.
Even as these Industrial Internet of Things (IIoT) capabilities become more commonplace, not every developer or project is ready to commit to wholesale edge automation. By architecting a control solution properly, end users can implement classic functionality today, while future-proofing their designs to take advantage of the edge moving forward (figure 1).
Developers unfamiliar with edge controllers may feel compelled to stick with tried-and-true control technologies and products. But the immediate impacts of ongoing labor and component shortages demand modernized systems that counter these challenges by taking advantage of new ways of accessing and using data. Developers and end users can embrace making their automation designs IIoT ready, and rest assured the required programming will build on past concepts.
When discrete automation operational technology (OT) originally shifted from hardwired systems created with electromechanical relay and timer devices to electronically programmable devices like PLCs, the Ladder Diagram (LD) language was created to help transition developers and technicians to the new solution. Visually, LD programming looks like the wiring diagrams used with hardwired relay and timer systems—but it greatly simplified programming, debugging, and modifying sequencing logic with a PLC.
As PLCs gained power and began to perform more complex functions like mathematics, analog loop control, and motion control, the associated languages had to evolve for performing complex expression processing, advanced process control, and motion. Languages like Structured Text (ST) arose to meet the new programming and debugging demands, yet LD was still available.
The development of PACs represented an expansion of PLC functionality to incorporate even more advanced application processing and communication functionality. PACs began to take on tasks we would associate with the edge today, although their dedicated real-time operating systems (RTOS) imposed some limitations. In many cases, users found it necessary to create complex algorithms, like machine learning strategies, using modern information technology (IT)–type languages like C++ and Python, running on PCs and industrial PCs (IPCs) working in conjunction with PACs.
Now edge controllers are on the scene, combining industrial-grade OT control with IT-type computing in one compact form factor tough enough to survive edge locations. Programming is evolving again.
Edge programming concepts
Some end users may opt to create edge solutions using a variety of hardware platforms and programming methods. This ad hoc approach introduces risk and requires time-consuming testing and integration efforts that generate late-stage changes and potential cost increases unacceptable in today’s industrial projects.
Selecting hardware and software offered as part of a wide portfolio of coordinated industrial automation products from a single provider ensures compatibility. A compatible portfolio of PLCs, PACs, edge controllers, and the software they run makes it possible for users to learn just a few programming tools and apply them to many products. Harmonized library objects lead to greater consistency and rapid development; proven code is readily reused on projects; and data is more readily transferred among systems.
To ensure future-proof solutions, users should look for product portfolios that embrace open programming, software, and communication standards wherever possible. Even open standards may benefit from industrial-centric extensions where appropriate. For an edge controller, this requires:
- deterministic control engines that support well-known and stable IEC-standard and C languages
- variable programming, with rich standard and user-defined data types
- extensive standard and user-defined function blocks
- support for object-oriented communications protocols like OPC UA that can seamlessly transmit data in context between deterministic and analytic applications
- support for OT-centric communication protocols like Modbus, Profibus, and Profinet
- Linux-based OS for the general-purpose computing part of the edge controller—bootable and upgradable independently of the deterministic control portion. Linux is lightweight and has high performance while using fewer CPU and RAM resources than other alternatives
- ability to run open-source apps like Node-RED and Grafana, as well as commercial and custom apps
- programmable using C/C++, Python, and other modern languages suitable for applications like machine learning and artificial intelligence
- support for IT-centric communication protocols like MQTT and HTTPS.
A true edge controller with these features is the best choice for implementing traditional deterministic control today and taking advantage of the edge computing evolution now and in the future.
Edge control and computing
Edge controllers are touted by many suppliers, but these devices vary significantly in performance and internal operation. For performance and security reasons, the most capable implementations use hardware virtualization to ensure the deterministic controller runs independently of the edge compute activities. Rigorous segregation of OT and IT environments is provided at hardware and software levels.
Today’s controls designers may be hesitant to apply new edge controllers to equipment and systems usually automated with PLCs or PACs. They may be concerned about the skillset required to make an edge control solution work reliably. They may feel unsure of the benefits gained from trying a new architecture.
These reservations disappear if the deterministic portion of the edge controller runs the same kernel as PLCs and PACs from the vendor and uses the same integrated development environment toolchain and toolchest of library functions. An edge controller can be used as a PLC or PAC today, even to the extent of using the same logic libraries developed for PLC/PAC members of the same product family. There is great power in using familiar methods tied to new capabilities to obtain new results.
If the edge compute portion of the edge controllers uses an open standard OS like Linux, users are free, in the future, to add almost any type of application needed. The immediate benefits are more compelling if the edge controller is available with a curated set of open source and commercial applications suiting the most common industrial computing needs (figure 2).
Control. The value of using an edge controller as a supercharged PLC/PAC is that the edge controller is natively ready to run visualization, data connectivity, analytical applications, and much more whenever the user is ready to take advantage of these functions. No additional hardware is needed; no cutovers or physical tie-ins are needed; and there is no requirement to interrupt execution of the automation application. Modern languages can be used where and when they are best suited for an application. End users gain the flexibility to extend their applications with edge capabilities when they are ready.
Computing. On the other hand, it is common for manufacturing facilities to have significant quantities of legacy automated equipment already in service. End users at these facilities may want to test the IIoT waters, but simply cannot jeopardize production.
In this case, using an edge controller only for the IT computing capabilities and OT connectivity, without even one line of control logic, is a practical approach. The edge controller can be installed anywhere on the factory floor, just like a PLC/PAC, and one edge controller can be networked to many PLCs/PACs, sensors, or other instruments to act as a gateway and information concentrator.
Operating in this capacity, users configure the edge controller to gather data, provide visualization hosting, and preprocess the data—such as filtering, averaging, or even executing local analytics—as needed, and to forward the information to higher level systems. Users can add and scale up this IIoT functionality at their own pace, without affecting any underlying automation systems.
Control and computing. Although the individual control and computing capabilities of edge controllers are important, users gain maximum advantage by combining them in support of one application. The deterministic “inner loop” directly accesses field-sourced data and controls devices. The edge compute “outer loop” combines deterministic data and real-time information from outside sources or analytics to advise the inner loop on operating parameters and tuning to achieve optimal efficiency (figure 3).
Edge controller drone application
Exemplifying the new potential made available by edge controllers, an end user recently explored using commercial drones for inspecting remote pipeline. The user was able to quickly incorporate the standard Linux software development kit (SDK) for a popular commercial drone into the edge compute portion of the edge controller. Using standard OPC UA communications, the company activated the drone to fly predetermined inspection routes either on default time intervals or on alarm conditions as commanded by the deterministic control portion of the edge controller. The edge compute portion captured video and telemetry data and then posted it to a secure web dashboard for remote operators to assess in real time.
Edge programming is more evolution than revolution. Originally, PLCs performed specific jobs and were programmed using correspondingly specialized methods. As PLCs improved and progressed to become PACs, the programming means and methods preserved the best aspects of the original Ladder Logic roots, but gained new languages, function blocks, and organizational methods as needed in support of added capabilities.
Edge controllers represent an exponential capability increase over traditional PLCs and PACs, so it makes sense that even more languages and applications are being added. This does not mean end users need to discard their experience or embrace all the added capabilities at once.
Instead, well-designed edge controllers remain highly compatible with existing PLC and PAC ecosystems, so users can continue to build on their OT knowledge base. When they are ready to extend applications into the IT realm, the same edge controller provides a general-purpose computing platform, effectively future-proofing their applications.All figures courtesy of Emerson
We want to hear from you! Please send us your comments and questions about this topic to InTechmagazine@isa.org.