Five things to know before diving into edge technologies

Edge computing can help companies gather and process data more quickly, but those getting started need to understand how much data they’re processing and how their systems work. Five tips for integrating edge computing are highlighted.

By Alan Raveling March 1, 2023

Image courtesy: Brett Sayles

Learning Objectives

Understand how edge computing can help companies realize Smart Manufacturing goals.
Learn where companies can start in their edge journey and understand potential challenges such as data storage, physical limitations and system integration.

Edge computing technology insights

Edge technologies help manufacturers get the most out of a Smart Manufacturing initiative.
Companies using edge computing still need to integrate the right technologies and systems to gather and process the data so the people responsible can make the right decisions.

One of the cornerstones of the Smart Manufacturing or Industry 4.0 paradigms is the application of edge technologies. The “edge” refers to the location within the controls process at or near the machinery or equipment. Edge computing located at the edge of the process can quickly process data with less communication latency.

It may be impossible for machine events that occur every second to be sent to the cloud, analyzed, and a response sent back in time for the system to act upon the data before the next event occurs. Placing the technology at the edge enables rapid data processing. However, it also requires proper implementation and integration.

Companies preparing to take the next step with edge technologies should look at these five ways to prepare for and integrate edge computing.

1. Operational challenges or opportunities for edge computing

Before jumping into an edge solution, an organization should define the objectives and results they expect from introducing it to the controls environment. Don’t be afraid to challenge perspective vendors to provide successful case studies and examples of their solutions working for clients in a similar environment. Understand the total cost of ownership (TCO) in years one, two and five, and spend time with the operations team to calculate the real savings or value resulting from the proposed solution. Not meeting expectations on the first few edge projects can sour an organization’s willingness to continue bringing in new technologies and solutions which can help them meet business objectives.

2. Data availability for edge computing

With a specific use case in mind, it’s essential to take the time to determine where the platform receives data and how that data makes its way to the edge solution. In instances where programmable logic controllers (PLCs), sensors and other equipment is Ethernet-based, it may be straightforward to have the equipment interact with the edge technologies. In locations where data resides in non-Ethernet networks such as RS485 or ControlNet (from ODVA), additional equipment such as input/output (I/O) gateways may need to be integrated with edge computing to enable access to the data. The polling speed of edge technology also must be assessed to ensure the process equipment can tolerate the additional load and the frequency of the information generated or collected by the equipment aligns to prevent unnecessary polling.

3. Data storage for edge computing

When creating a solution that leverages edge computing, application developers must determine how much controls data must persist and where the data can exist. In applications such as machine learning (ML), where large volumes of data are necessary for training, edge technologies may not be able to hold the data. It would need to be held upstream, such as in a data center or a cloud solution. Large data files such as images or videos fill up storage and create a larger strain on networks for data transportation if there’s a need to archive or send it to an alternative location. Determining each piece of data’s lifespan and where it resides at each lifecycle stage helps aid in the design of edge solutions to help mandate which technology to use.

4. Physical considerations for edge computing

One often overlooked area when discussing edge solutions is the real-world issues with installing equipment. Will the equipment be installed within machine control panels or in a new or dedicated panel? Is there a standard receptacle free for the power cord, or will the edge device need to operate on 24 V power? Is networking accessible to the edge devices, or will new networking cables or equipment be required? Project plans can be thrown off schedule without evaluating the physical locations where edge solutions are slated to be installed. Once installed, those supporting the applications and solutions should understand the access to the edge technologies. They may not have the appropriate certifications or safety training to access the panels or locations of these platforms.

5. Developing technical skills for edge computing

Many edge solutions use standard operating systems and programs underneath impressive-looking web-based management portals and dashboards. Organizations should seek to understand from vendors what skillsets and activities are involved in the ongoing maintenance and support of edge-computing technology. For example, the use of Linux instead of Microsoft Windows on many of the newer edge computing solutions has resulted in a skills gap in areas where most applications have operated on Windows-based operating systems.

Now, edge technologies are more frequently using microservices such as those from Docker or Kubernetes, which operate on concepts and architectures very different from the applications and platforms that operate process control solutions. Understanding who and how any support, maintenance, or changes to edge devices can be carried out is critical to ensuring the sustainability of any implemented solutions.

Controls engineers, edge computing collaboration

Edge technologies offer various capabilities and possibilities for organizations bringing excitement and concern. Without proper research, planning, and a solid implementation strategy, the first project could also become the last. It’s crucial for controls engineers, data analysts, business leadership and the vendor to work together to understand the best set of technologies for the specific environment and the requirements and prerequisites needed for success. Be wary, but remain curious, of those who promise drop-in solutions with minimal effort required for implementation. This landscape rapidly changes, and it is important to stay informed of the new capabilities and functionality available in edge solutions. What was considered impracticable a few years ago is now a commodity. For organizations wanting to integrate edge solutions into their environments, start small with challenges that can bring quick wins, then apply those learnings to larger and more complex challenges. The possibilities these technologies bring are exciting and will transform how engineers think about machine and process control.

Alan Raveling is OT architect at Interstates and a 2021 Engineering Leader Under 40 winner. Interstates is a CFE Media and Technology content partner. Edited by Chris Vavra, web content manager, Control Engineering, CFE Media and Technology, cvavra@cfemedia.com.