The ongoing quest for deeper visibility into operations is driving the adoption of more network architectures than previously used in industry.
In building an Internet of “Things” (with many of those “things” being brownfield assets) we create numerous network devices. And when a “thing” becomes a network device, it picks up new roles and responsibilities.
Firstly, our “thing” now needs to be configured and managed. Additionally, our “thing” needs to have enough intelligence to know what data to publish, and when to publish it. This all means our “thing” needs to make some autonomous decisions based on prescribed business logic.
And that’s not how we handle things in the world of automation.
In automation, there is a master or control device calling the shots and we don’t expect to find much in the way of independent thinking from devices at the very edge. But what’s appropriate for automation and control doesn’t always fit the world of monitoring and analytics in the industrial IoT. Why is this? In automation, the data consumer is a control device that is typically located within a cable-reach of the edge device. Data is requested, consumed and discarded locally and in real time.
The current trend is to push more and more intelligence out to the edge of the network. Sometimes that edge is a device gateway. In other cases it can be a smart, intelligent sensor. Since the edge typically lays well beyond the IT closet, or even beyond the control panel, those edge devices for the IoT may have to be hardened and ruggedized to meet the requirements of the environment they will live in.
Here’s the takeaway: technology is in our favor.
It costs relatively little to put processors, memory and network stacks in small devices today. But it’s worth recognizing that these IoT architectures require not only new types of intelligent, industrial devices, but also new skills. This trend will only grow as networking and IT technologies continue to intersect with operations.
