why? Image processing is cheap (laptop), cameras are cheap, this would turn manual P&P into monkey on a tricycle/video game type of affair, where you can put totally clueless person behind the desk and ask them to do what the pretty green lines on the screen tell them to.
a few thoughts of why not:
- a manual PnP machine will become quite large with a laptop , the screen of the laptop has to be in view with the attached keyboard infront not very ergonomical, an external large monitor will take even more space.
- you have to boot the laptop for 30 seconds, run the correct program, be annoyed all the time with all kinds of OS bullshit popup messages like updates/virusscans etc.
- there are a lot of problems reported to get two identical USB camera's on one PC working
- you probably have to write your own software which will take weeks/months to get correct, only interesting if you go to a DIY fully automated PnP
you are effing with me, right? complaining about how difficult and unwieldy laptop is to operate?
especially compared to setting up full blown P&P
Consider the benefits of manual, but vision augmented p&p - you put absolute MONKEY in front of your manual pick and place machine, and all that untrained monkey has to do is follow pretty pictures on the screen. This means
you dont have to do the boards, intern/your kid/cheap manual labour.
Most expensive part of P&P machine is ... machine itself
precise repeatable quick mechanical bits. Software is cheap in comparison.
AR program wouldnt even need to know how to read board files, it could learn on the first pass - make first board yourself, and on the next pcb program knows where all the components go and overlays accurate positions/orientation/directions.
hmm those guys could give it a go (granted it should work better than their slow clunky demos)