BLOG by Justin Hogue, PE
Tuesday, October 20, 2020
Press Contact: Liz Densmore, APR; Ldensmore@LECINC.com; 601.939.8535
What exactly is Edge Computing?
It really all depends on who you are asking...
For some, like IT and software development, edge computing means a new piece of hardware, or embedded PC, where certain data values are calculated based on I/O or sensor status, which is then used to calculate process-related or predictive analytics.
Many of these cloud-based software solutions are strictly focused on concentrating the data collected for analytical purposes, and rarely provide any real degree of “smarts” in the field.
So, from that perspective, the idea of converting sensor or I/O data into actionable intelligence before the data is transferred to the cloud is an additional step, and amounts to placing logic processing and automated decision-making remote to the application server at the equipment or process site, a.k.a. “the edge,” where the connectivity and data visualization literally meet the process.
Another element at the edge that is often overlooked is edge security.
Consider the “edge” as the doorway into a control system.
It makes no difference how powerful or strong that door is if there is no lock to keep the door from opening up to anyone who turns the knob.
That edge device is essentially acting as the "bouncer," only allowing those who have the appropriate access through.
Without the "bouncer,” that control system is vulnerable to all sorts of chaos that could cripple a system before you realize anything even happened.
For those in industrial manufacturing and machine applications, like LEC, all of this is nothing new - we’ve spent the last 30+ years working every day at “the edge.”
However, the majority of the solutions we have implemented have been historically, and still are, mostly intended for local use in monitoring process variables in order to better manage production efficiency, raw materials, and quality, while providing corporate-level data for a multitude of purposes.
But, that doesn’t mean we can’t utilize these existing automation and process control systems to provide a highly robust platform for additional computing at the edge.
Another question to take into consideration is: Why would someone add another processing platform…and additional complexity…to an existing industrial control system that already provides a means, or platform, capable of mathematically manipulating I/O or sensor data for virtually any purpose?
We don’t serve up graphics or host User Interfaces at the edge, so why would we need more computing horsepower to transform machine performance data than what is needed to automate at a process manufacturing plant?
LEC has been doing edge computing in industrial applications for a very long time.
When these capabilities are leveraged in combination with modern IoT technologies - connectivity and cloud applications - the possibilities are just about limitless.
With industrial applications involving machine control, the ability to process logic and make decisions cannot be subjected to the latencies and potential inability to communicate reliably.
Safety, security, and economics are the top concerns, and when industrial automation experience can be combined with cloud application and communication experience, most any controls system can be dramatically improved.
System architectures that allow processing power to be placed where it can be most efficiently integrated at different levels or locations within the architecture will best cover the primary concerns mentioned, while enhancing scalability.
Send us your thoughts at: info@LECINC.com.