The value of real-time machine data

Editor in Chief Joe Feeley disputes the value of the IT model that says you help your customer more by spending dollars on data manipulation capabilities rather than on improved machine performance.

 By Joe Feeley, Editor in Chief


et me tell you about an unexpected conversation

I found myself in with some IT professionals about the value of real-time data. This was at a holiday gathering where the only serious conversations normally came from cautiously figuring out what that funny-looking finger food might actually be.

We hear about and participate in a lot of talk about how hard it is to make all the right connections so that real-time factory-floor operating data in your machines is of optimal use to the enterprise. Ah, yes, the “enterprise.” Many of you, probably most of you, don’t have that requirement inside your own company’s walls, but a growing number of your customers do and you’re expected to have a lot of the answers to satisfy their demands.

I’ve been a little cynical about this subject for a while now, so I took the opportunity to ask just my contacts what their IT groups were really, I mean really, doing with the machine data, and more importantly what the company managers were doing, I mean really doing, with the data to better run the company, err, enterprise.

I know the appropriate technology is available—and it’s getting easier to use all the time, although that’s mostly when newer control technologies are involved. I’m also not impugning the value of real-time and/or deterministic machine control to improve product quality and yield. That’s becoming more and more evident every day. It’s what happens to the data when it leaves the factory floor that I grumble aloud about.

One of the IT-ers works for a manufacturing company, the other for a consulting business with a few manufacturing clients, which meant our conversation wasn’t just a theoretical exercise while enjoying a cold beer.

So, I asked them to explain why a company needs real-time machine operation data in the office? I’ll buy every hour, I said. Every shift might be stretching it in some cases. But, every millisecond? Come on, who are you trying to kid? In fact, I added, regardless of the scan rate, what are they using it for?

Their rebuttals were as abundant as they were adamant about the way the data can transform companies’ response and planning culture. They also were a little uncomfortable defending themselves against such unbridled heresy that dared question what good the data was doing in their own organization.

Then one of them let it slip. It actually hasn’t lived up to its potential yet because management is still wrestling with more important issues. So, she said, they don’t yet know how to use very much of the data, but they feel they need to have it.

Ah-hah. They’re too busy doing the important stuff? Is that what she said? So, this isn’t the important stuff?

This is what makes me crazy. Are too many companies prematurely busting their system integrators and machine builders to raise the data manipulation capabilities of the machines and the software expertise of the controls support staff? That comes at a price, and it’s a tough spot to be in when an industrial OEM has only so much project development money and help around.

I think you can be doing more with those limited development dollars to keep pushing the limits of machine performance, and that’s how you can best help those customers today. How do you tell them that?

The day surely will come when the data from just about every factory machine will cumulatively help companies make better and more-far-reaching decisions. I retain my heretical beliefs that it’s not helping machine builders help very many customers today.

I’m prepared for you to burn my thoughts at the stake if I’m wrong here. Am I?