How often do you hear, “The vision system’s stopped working!”?
Too often, I’m sure. Yet you and I both know that vision systems don’t just stop working, (unless the power is off or a component has failed – all rare events.) So what’s going on?
Well, if the system really isn’t functioning as it’s supposed to, more often than not it’s because something has changed. Either the part is different in some way – perhaps the surface is dirty – or the image has changed due to dirt in the system, or a hardware component has been moved. But providing a system is maintained in its original state, it won’t just stop working.
So I was a little puzzled by the tone of “Top 10 Machine Vision Improvement Opportunities” (Vision & sensors, June 30th, 2009.) This article seems to imply that you can make your systems perform better by working through their ‘Top 10’ list.
The problem I have with this is that eight out ten items are things that should be addressed during system design; using the correct lighting, for example. If the lighting is wrong the system was never designed correctly from the outset. (By the way, the two I agree with are calibration and stability.)
Now I don’t dispute that there are ways to improve the performance of vision systems already installed, I’m just suggesting that this list goes after design issues rather than maintenance issues.
If you want to read my thoughts on machine vision maintenance issues, check back in a day or two.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment