What is an “edge” device in terms of network architecture for today’s Industrial Internet of Things? Classical networking practice has had it’s definition. But how do you extend the definition in today’s industrial networks with perhaps thousands of devices at the edge? Do you label all those smart devices as edge?
I have been spending much time with Dell Technologies and its IoT division. It has built a computing device with a multitude of connection ports, data storage, and computing capability. This device is named Gateway, but it is labeled as an edge device. Meanwhile I interviewed two GE Automation and Controls executives who labeled controllers (PLCs) as edge devices.
I ran across this article by ARC Advisory Group’s Greg Gorbach. I’ve quoted some of it below. You can read it in its entirety here. He analyzes a number of points of view. Does it all matter to you what is called an edge device? How do you configure a modern IIoT network?
Power of Edge – Greg Gorbach
What is the industrial edge, and why does it matter? Is it network infrastructure? Can the edge be found in a sensor that feeds a controller in a plant? Or is it in a smart machine that’s in service halfway around the globe?
In networking, an edge device is a device which provides an entry point into enterprise or service provider core networks. Examples include routers, routing switches, integrated access devices, multiplexers, and a variety of local area network (LAN) and wide area network (WAN) access devices. Edge devices also provide connections into carrier and service provider networks. Network providers and others have been pushing intelligence – compute power and the ability to run applications and analytics – to these edge devices for some time.
But the growth of the Industrial Internet of Things (IIoT) extends the ‘edge’ beyond the network devices, into industrial and commercial devices, machines, and sensors which connect to the network. Edge computing and analytics can, often should be, and increasingly is close to the machines and data sources. As the digitization of industrial systems proceeds, we expect that analysis, decision-making, and control will be physically distributed among edge devices, the network, the cloud, and connected systems, as appropriate.
These functions will end up where it makes most sense for them to be.
IIoT will change the way industrial organizations generate, collect, and analyze data. Data will be generated faster and in greater volume than ever before. This will require today’s plant information infrastructure to evolve. One part of this new infrastructure will be intelligent edge devices, which will include the latest generation of controllers, such as DCS’s, PLC’s and PACs. Besides providing control, these edge devices will securely collect, aggregate, filter, and relay data, leveraging their close proximity to industrial processes or production assets. They will also be capable of collaborating with powerful analytics tools, detecting anomalies in real time, and raising alarms so that operators can take appropriate actions.
With edge computing and analytics, data is processed near the source, in sensors, controllers, machines, gateways, and the like. These systems may not send all data back to the cloud, but the data can be used to inform local machine behaviors as it is filtered and integrated. The edge systems may decide what gets sent, where it gets sent and when it gets sent.
Placing intelligence at the edge helps address problems often encountered in industrial settings, such as oil rigs, mines, chemical plants, and factories. These include low bandwidth, low latency, and the perceived need to keep mission critical data on site to protect IP.
As you think about digitizing and transforming your industrial operations or your products and services, pay special attention to the edge. Consider the optimal location for analysis, decision-making, and control, and the best way to distribute these among edge devices, the network, the cloud, and other connected systems.