The technology is not only expensive, but also difficult and time consuming to integrate. The normal maintenance crew tasked with keeping the machines running may have some programming expertise, but that isn’t nearly enough.
It takes a machine builder or a hired gun (programmer) to come in and spend a few months hacking code, implementing the system, and debugging it. We all know that the debug time is where the real time and frustrations come in. No offense to the maintenance crew, but this is above their pay grade.
This is what I hear a lot when I talk to customers about their control problems. Why not industrial RFID? If RFID is truly as “bad” as some people make it sound, then why do so many engineers use it? It comes down to this:
RFID can provide huge returns on investment—there are situations where there are simply no other or no better options than an industrial RFID system. For example, let’s take a production line with 10 independent stations. These stations contain 10 independent control systems including motors, drives, inputs, outputs, and robots. These systems don’t work together and they may even be made by different companies to specialize in specific tasks. One company may make a best-fit system for door installation and alignment while another one specializes in paint booths.
The different stations in this example require completely different technologies, different know-hows, and are never handled by the same company. They do need to somehow communicate with one another (not really, but both need access to the same or at least similar data). For instance, the doors must not be installed unless the body is painted properly.
The simple solution is to use RFID. Once the operation at each station has been performed properly, a simple pass/fail bit can be written to the tag. At the end of the line, you would then have a series of 10 pass/fail bits correlating to the 10 independent stations. A reader at the end then reads the tag to verify that all stations performed their tasks correctly. If any station failed, the part could be sent to rework.
In this simple case, what would you do without RFID? You could store the information in a separate control system that connects to all of the stations. This, however, takes special programming at each station and programming at the management system. If a PLC is doing the upper-level communication, where is the product pass/fail data stored? In the PLC? Not likely. Batteries die, processors fail, and then what? What will be the fate of all parts currently being produced on the production line? You could instead store all the information on a server that is backed up regularly. Sounds like you need to get IT involved. Both factory floor PLC programmers and computer programmers have different knowledge bases and their own agenda and may have a difficult time working together.
Every RFID tag, on the other hand, is a little database. This database doesn’t have to be backed up; there is no battery to replace. The data retention on these tags is 10 years—far longer than it takes to produce a part—and IT never has to be called. Communicating with the tag is not all that hard. Built-in function blocks in the PLC easily read from a tag when a part pulls into a station and then write back to the tag when the task is complete. The read is typically done first because there may be cases where a station task can’t be completed unless all previous stations performed their tasks correctly.
So the bottom line is, for factory floor parts tracking, industrial RFID probably isn’t a bad idea. A machine builder who is used to PLC programming will have no problem interfacing with an RFID system. No additional high-level programming skills are required. RFID hardware is certainly more expensive than an 18 mm inductive sensor, but a lot cheaper than regularly scrapped parts or repeated warranty claims due to faulty products leaving the plant. Is integration more difficult than with an inductive sensor? … Yes, but just like anything else, a bit of preparation helps—and it never hurts to ask your supplier for some input and help—it certainly makes the extra integration time manageable. It could be worth it. Also, after you do it once, everything becomes easier. You can just copy and paste the ladder logic code for use on the next job. The next time you think tracking of any sort is needed, give RFID full consideration.