Tuesday, December 29, 2009

Prediction number 2

This concerns interface standards and particularly the adoption of GigE. I believe there’s growing momentum behind GigE and that by the end of 2010 this will have overtaken 1394(b) in popularity. My reasoning on this is, as with prediction number 1, driven by the demand for higher camera resolutions, and hence bigger images.

This is leading to bandwidth problems that need 1394(b), CameraLink or GigE. Unfortunately 1394(b) has a reputation for being somewhat “cranky” while CameraLink is a little intimidating for the average machine vision user (all that framegrabber stuff leaves engineers looking for an easier ‘plug ‘n’ play interface.) This leaves us with GigE.

The thing with GigE is that once people have tried it, they seem to like it and don’t want to go back. This, I predict, will result in growing momentum and adoption will accelerate accordingly.

You might also notice that I’ve left out any discussion of cable length limitations. Contrary to what the camera manufacturers tell us, I believe that for most users this is a non-issue. The number of situations where we need the PC to be more than ten meters from the camera is actually fairly few so cable length isn’t a high priority.

1 comment:

Anonymous said...

I don't agree with the cable length issue. When building robot cells with "built-in" machine vision, 10 meters is usually just too short.