And who might these dinosaurs be, you ask?
I've discovered they are in our midst and in places where you might not think they are. They are buried in the control systems of OEMs-past. They are in conveyance systems whose companies have since disappeared.
They are in processes where only one person knows enough about the system to keep it running.
How can we operate like this?
The old adage of forgetting to drain the swamp while you fight the alligators comes to mind. If it's not broke, don't fix it--that's another one.
These old, tired, and proprietary control systems simply get forgotten until something happens--usually, they break. The processes they once controlled infallibly have come to a grinding halt or simply don't work as well as they used to.
"Sorry, but Roy doesn't work here anymore, and he was the only guy who knew those systems," says the tech support person from the machine builder. "We haven't made those for more than 10 years. But we do have a replacement system that we can integrate for you." Sound familiar?
I had a phone call a while back from a friend who had driven from Canada to Ohio to help a customer whose process went for a loop (no pun intended). My friend still worked for the OEM where he originally wrote the software, and the platform was an Allen Bradley PLC-3/10 processor. He was unfamiliar with the system and it had been a few years since he had anything to do with hardware setup.
But he was the man, so off he went. Seems that a card that interfaces with the remote I/O in the system had blown. Piece of cake to replace, right?
Three days later it's still not running. He called Allen-Bradley support and discovered there wasn't anyone there that could support the PLC 3/10. The PLC-5, as most of us in the A-B sphere of influence know, replaced the PLC-3 as the processor of choice in 1988.
The gurus of the past had moved on.
A-B's tech support is not at fault here. This situation evolved because the owner of the hardware couldn't explain himself properly because of his unfamiliarity, as well as the lack of certified PLC-3 guys on A-B's end.
So when he called, I asked him if they had configured the new module by adding the I/O chassis logical rack list. I think his response was "What?" Problem solved. Fundamental, but not surprising.
I have also dealt with a customer in Ontario who still uses product from the 1970s. Scary, to say the least.
It isn't the legacy part of this that's important. It's how are you going to handle the transition when it happens. And it will happen.
Do you accept the new and improved proprietary solution from the OEM to replace the existing system? Do you scour the Earth for someone who can spell PLC-3/10? And what happens when the beer truck theory plays out? That's when the guy you've finally found can help you gets hit by a beer truck.
The project that I had been working on and had described in the past two columns took a twist. I discovered that the reason they wanted to do some of the things they wanted to do was to prevent the same issues that I described above.
They don't want to be caught with their proverbials around their ankles. What this really meant was that they had two systems running in their facility and the only person available to fix the problems was from the OEM.
Closed-system architecture and software. Can you say expensive maintenance contract? And to boot, the service guy has since left the company and is supporting these legacy systems on his own. Seems he has enough work to keep him busy, and no amount of money necessarily gets him anywhere, since he has his choice.
Great work if you can get it, but this spells doom for any OEM who subscribes to this mentality. Also spells trouble for the customer using that OEM.
Open systems have a great place in the industry, as do the people who support them. It would be very prudent to plan for the future now, and not run into the surprises that have showed up in the very recent past.
Find those dinosaurs and kill them. If your only point of support equals one, then you know you're in trouble.
Jeremy can be contacted through his web site:
http://www.tsuonline.com.
