Increasing Edge Expectations
As edge computing grows in scope, it can be difficult to keep up with the latest definitions. This article breaks down the components of modern edge applications and what they need to succeed in modern industrial use cases.
In the modern era of industrial automation, there has always been an “edge” where the physical, real world transitions into digital systems. Industry 3.0 involved the application of digital monitoring and control, while Industry 4.0 is marked by the expansion of connectivity and the Industrial Internet of Things (MoT). The edge has existed throughout this progress, but it has grown in both size and capability to meet end user needs, especially in the area of data management and analysis.
These developments bring new possibilities for end users everywhere. In fact, there are so many opportunities it can be hard to know where to start, or to realize the extent to which barriers have fallen. Instead of just performing older tasks better, advances in edge connectivity and real-time control expand the horizons of what is possible. Users can start with certain practical goals, but they should also actively consider new and perhaps even unconventional approaches.
It is important for end users, original equipment manufacturers (OEMs), and systems integrators (Sis) to familiarize themselves with the necessary features of a comprehensive edge platform. The right edge platform will remove roadblocks to creating innovative applications, future proof installations, and provide a host of other benefits.
While the “edge” has existed since the beginning of industrial digitalization, a modern and comprehensive edge platform can create truly innovative solutions.
Edge platform basics
Edge applications bridge the gap between operational technology (OT) industrial systems and information technology (IT) resources. Organizations must be able to rely on OT implementations to run for years with minimal upkeep while effectively connecting with IT.
Part of the attraction of edge applications is the freedom users experience in choosing from an entire marketplace of hardware, software, and technologies to create a solution tailored to their needs.
However, this extreme openness is more demanding than it might seem a first glance. A full-custom approach means the developer is responsible for ensuring the capability, performance, compatibility, cyber-security, and supportability of all elements, in conjunction with each other.
Industrial applications call for greater assurances on each of these issues. Users will find that a complete edge platform from an experienced industrial automation supplier addresses each of these points by providing a coordinated, compatible, and wide-ranging portfolio that includes:
- scalable hardware
- flexible software
- extensive connectivity
- comprehensive cybersecurity
- mobile accessibility.
When developing an application, designers quickly discover there is no “one size fits all” when it comes to edge hardware. Not too many years ago, a traditional programmable logic controller (PLC) was a primary digital interface method for connecting with most field signals. Since then, field devices have become more intelligent, and some are directly MoT capable, while real-time control and computing options have also multiplied. PLCs in small, medium, and large form factors remain relevant, although the larger versions are now often referred to as programmable automation controllers (PACs) because of their more extensive capabilities.
A key change at the edge has been the introduction of edge controllers. The best implementations of edge controllers perform real-time deterministic control just like a PLC- but add the onboard, general-purpose computing necessary to support MoT applications. The most demanding applications may call for a full-fledged industrial PC (IPC).
Edge applications can be diverse to meet specific needs. They can range from gathering PLC or intelligent sensor data, to performing more complex aggregation and computational roles. Therefore, edge software must flexibly meet the need for connectivity, computing, visualization, control, and analytics, which are all different but related. Software also should follow industry standards, while being interoperable, open, and extensible.
Edge software must operate locally, while seamlessly connecting with higher-level on site or cloud-based resources. It may be called upon to gather field data, perform preprocessing, support local visualization, communicate information to other systems, execute local control, and carry out analytics. This is a tall order, and the software also must be easy to use, enabling end users to develop and deploy applications.
End users have a mix of control and edge application needs, but each solution relies on connectivity, which involves both hardware and software. From a hardware standpoint, connectivity includes hardwired input/output signals and serial communications, along with support for industrial fieldbus and wired/wireless Ethernet protocols. Software provides the appropriate communications protocols, which need to be native to the edge platform, with the option to add support for more as needed. Complicating connectivity is the fact that both OTand IT have specific and different protocols unique to their implementations, so any edge platform must readily work with both.
Great connectivity demands equally capable cybersecurity. Edge platforms by their nature make OT equipment and operations potentially accessible to cyberattackers, who may steal or alter data, or cause improper system operations. Creating an edge solution using unrelated products from various suppliers means the developer needs to ensure not only the functionality and interoperability, but also the overall cybersecurity provisions. Instead, there is a case to be made for selecting products from a portfolio of industrial-specific options designed to comply with cybersecurity standards and best practices.
Although there are many examples of edge solutions not needing a specific operator interface, it has become common for the edge to support local, remote, mobile, and browser-based visualization of direct field data and analytical results. Some applications may need all forms of these interfaces, supplied via industry-specific software and human-machine interfaces (HMIs), as well as more general open-source tools. A complete edge portfolio should support all these options.
For end users, OEMs, and Sis, most projects can succeed with the right mix of hardware, software, connectivity, cybersecurity, and mobility. Connecting with disparate data sources, performing computations, and communicating and visualizing the results are each unique but related tasks. Following are a few examples where designers creatively pursued comprehensive solutions to meet their goals.
An OEM provides connectivity and monitoring solutions specialized for supporting remote sites, such as oil wells. The OEM wanted to offer advanced capabilities-such as flare monitoring, adherence to security practices for personnel, and even vehicle license recognition – but this required a level of computing power that was previously impractical for field installation. However, by using an IPC and associated industrial-grade edge software, it built the exact edge computing solution needed to process data locally, which reduced the bandwidth needed for upstream communications to higher-level systems.
Another organization needed a way to monitor the health of Bitcoin server farms spanning multiple sites. It needed a simple way to monitor existing systems, add new sensors in parallel to existing devices, and enable remote access for a centralized view of all operations. Edge-capable hardware and software made it straightforward to add these robust capabilities to systems formerly limited to largely proprietary HVAC control and monitoring systems. The solution can incorporate external data, like weather, and perform analytics spanning many sites.
A third example is a manufacturing company that needed a better method for weld inspection. It developed a machine-learning algorithm that could be run for each weld to compare the actual results with what was expected or predicted by the model. Suspect situations could be indicated to users, and the system could even put a hold on the operation until the issue was fixed. Edge solution hardware and software selected from a comprehensive portfolio let the company create the system it needed, while minimizing concerns with respect to processing power, compatibility, and cybersecurity.
Expecting more from the edge
While the industrial edge has existed in the digital age, recent technological advances mean much more is possible at the edge than ever before, and by a large margin. This also suggests that implementers can think bigger and more openly than they are accustomed to by considering a wider range of options.
Silvia Gonzalez is the director of product management, software, for Emerson’s machine automation solutions business. She is responsible for developing lloT, industrial automation, and controls technologies that bring increased value to customer operations. Gonzalez has a bachelor’s degree in electrical/electronic engineering from Universidad La Salle, Mexico, has received a digital business strategy certificate from MIT, and is based in Houston.