Diversity in machine vision

Diversity is a good thing, right? I don’t think anyone would argue that a diverse workforce is not an asset to a business. Isn’t it common sense that bringing together people of different backgrounds and experiences enriches an organization by nurturing innovation and creativity? Of course it is.

So why don’t we do the same with programmable logic controllers (plc’s)?

We could. We could use whatever best suited the specific application; Mitsubishi on this press line, Omron on that assembly station, Siemens in the warehouse, Allen Bradley in the forging cell. But what a support headache we’d give ourselves.

See what I mean? Sometimes homogeneity is a good thing. Yet so often when I tour a plant I see they’ve opted for diversity in machine vision. It’s not uncommon to see Keyence, Cognex InSight, DVT and Omron systems used alongside proprietary products that may be LabVIEW or Halcon-based, but why would you do this?

The answer is usually along the lines of it being the best fit for the application. That may be true but I think it’s a bad reason because it results in the maintenance department trying to support a host of systems they know very little about. So what should a company do?

Well my answer comes in two parts. First, for all those smart cameras that you install yourself I suggest you pick a major vendor and stay with them. I realize this means marrying Matrox, Cognex or NI, but come on, it’s not forever!

Second, what about when you buy manufacturing equipment and a vision system comes pre-installed? Does it matter who made the camera and software? Well I would argue that you should try very hard to limit variety. You don’t want to be keeping a dozen different cameras in the parts store, and you do want the ability to do first line maintenance (checking image quality, adjusting exposure time, that kind of thing,) so push your equipment builders to stay with the established players.

I realize this attitude is what makes life hard for new entrants to the machine vision business, but this is not an equal opportunities situation. Machine vision users need to be selfish and put their business first.