AI-Powered Electronic Component Sorter(github.com)
github.com
AI-Powered Electronic Component Sorter
https://github.com/Vanguard-s/Electronic-Component-Sorter
5 comments
> automatically identify computer motherboards from pictures
An extension of this would be to detect any visibly-altered components that might indicate a device failure or an unexpected change in the product being delivered.
Though you'd have to have really good pictures/angles to detect something like bulging electrolytic capacitors.
An extension of this would be to detect any visibly-altered components that might indicate a device failure or an unexpected change in the product being delivered.
Though you'd have to have really good pictures/angles to detect something like bulging electrolytic capacitors.
Call me when this project includes more robots. After several years and workshop upgrades my junk drawer consisting of odd sized bolts, circuit components, zip ties, random batteries etc is so heavy that I need a truck to move it. I would pay big monies for a simple device with a precision claw for pick and place robotics that can help sort such a mess. If it knows the difference between hardware that differs even in small details or dimensions, cool, but coarse like-with-like sorting that simply separates bolts from nuts from batteries would be useful as well, and even seems doable offline.
If this worked well you would imagine a demo consisting of more than a single through-hole component.
Doesn't Google already have an AI that can identify anything based on a picture?
Paired with something like the URL for the manual/documentation at the manufacturer's website and/or the Wayback machine (if the manufacturer is defunct and their website no longer exists), and something like Octopart (for ordering replacement components), it could have a lot of applications...
Sort of like the digital equivalent of an AI "what plant/flower/shrub/fruit/fungi is this?" for the natural world, but for man-made electronic boards and sub-components...