Edge devices in industrial settings range from tiny Raspberry Pis reading PLC data to large server racks running algorithms. When deploying them, consider three things: compute power needed, data volume and type, and security concerns.
If you can get away with kind of a minimal amount of data points, most likely you're going to be able to have those computing resources locally. And so looking at just on a basic level, what kind of horsepower, I'll call it, for the hardware that you're going to need to run that solution, and then the amount of data you'd be feeding it tied in with security concerns would be the 3 primary things I'm looking at for these types of solutions.
Okay, and as far as these types of solutions we're talking about as edge devices, you know, can you describe what an edge device is?
An edge device is going to be something that runs as close to the edge of the plant network as possible. So it could be something as small as a Raspberry Pi just pulling data from a PLC and then pushing it to a local database or doing some sort of light logic built into it. It could be a sensor gateway that I mentioned earlier with these different sensors attached to it, piping data directly to a cloud.
And then also could be a server.
It could be a large server rack that's running one of these fancy algorithms and hosting all of these different data streams and piping that to different places. So you have data visualization, kind of data transportation, and then also some data processing locally there.
