Morning Scott,
At the plant I work at, we use a pantone colour book as a quick reference to ensure the colour is accurate. Simple, but effective. :-)
Colour Book.docx

Try using some of the paint store apps, like Project Color from Home Depot. Benjamin Moore and Pittsburgh Paint also have similar products.

In the past I have created "working standards" to use at the end of the production line. I added these reference standards to my calibration system. You would also need to "color test" the associates. Also jd marhevko has a presentation addressing calibrating Humans.

Make sure you use the standard that is defined in your internal or customer's spec. They be using AMS-STD-595 color index system. Chips can be purchased, I think they are around $10.00. Lay the chip in the same plane as the surface your are examining. Light bounces in funny ways and if the chip is not in the same plane, the inspector may think the color is off when in fact it is true.

Pantone books could be used for verification. The pantone number could be included in the production paperwork then the pantone book could be used as a standard reference when performing your quality checks. Make sure whoever is performing checks is tested for color blindness.

Your solution might come from an old astronomy technique. Before computers, astronomers made discoveries by comparing two photographic plates of the same patch of sky, identifying something that shifted (like the planet Pluto, an asteroid, or a comet). To greatly simplify this task, they used something called a blink comparator (https://en.wikipedia.org/wiki/Blink_comparator). An 20th century amateur astronomer (and professional interior designer), Ben Mayer, invented a projection version that projected images to a large screen. Mayer invented his version as a low-cost but high-quality tool that amateur astronomers could use.

I could find no picture of Ben Mayer's projector, but the end effect could probably be accomplished with two projectors of the same model: one projecting the color standard, the other a live image of the paint. The image used as the color standard should be a photo taken of "good" paint at the same location on the factory floor so that the image had the same lighting, camera angle, texture, etc. It seems to me an overlap of the two images might serve as an indicator of the two colors matching: If no discernible color change in the middle, the paint would match the standard.

If you are interested, images of the pre-Ben Mayer blink comparator can be found by searching for images using the phrase "Blink Comparator Used to Discover Pluto."

I hope this helps.

Thx everyone for your responses so far. I should have been more clear - we already use most of the manual methods you have suggested. I am looking for something electronic that will read the color on the fly, compare it to the standard that is already in our system and give a pass/fail.

If you're looking for a more standardized non-visual based check I would look into BYK instruments. They have several portable meters that can measure color and/or gloss. Then it's just a matter of establishing your target values and acceptable tolerances. I don't know if their systems can be automated but I would suspect you could train someone to operate it, and it's a fairly quick test.

https://www.byk-instruments.com/en/Color/c/1977

We use a Nix Colour Pro Sensor. It links to an app on a phone or tablet. It's super quick and compares against reference sample you enter. What I love too is that it records all the values (date and time they were taken too). It looks like its about $400.
https://www.nixsensor.com/nix-pro/

The Nixsensor mentioned above looks like a good alternative. Just be aware that the sample you are comparing it to has to be in good condition. Any paint samples or chips should be stored in a protected area (i.e. envelope or container) that will protect it from light and fading. Or checked for accuracy against a control and replaced as needed. Definitely a less expensive option than a spectrophotometer.