Arduino Anderometer

Hello!

I'm investigating methods to confirm NVH in ball bearings, and I repeatedly lead myself back to anderometers.
Anderometers basically apply opposing axial loads on bearing inner & outer races via a pusher to exacerbate imperfections, rotate the inner race (typically @ 1,800 rpm) via a drive unit, and then use a velocity sensor applied to the stationary outer race to detect vibration that would lead to noise.
The only problem is that the cheapest anderometer I've found costs 6-figures, and I'm looking for a significantly less expensive solution.

I'm very much a n00b when it comes to electronics; the most I've done is just built my own PC, so keep expectations low.
Are there Arduino compatible linear (not angular) velocity sensors?
I'd like to have an adjustable pass/fail threshold with a simple visual indicator (ie. red light = bad, green light = good) for bad & good bearings, so if I have a bad bearing that this system would have passed, I can adjust this pass/fail threshold accordingly; if the system can maintain the "bad bearing red light" if one occurs until the test is reset, this would be even better.
Is this project feasible using the Arduino platform?

I'm looking for a turn key solution with high precision & easy for my end users to use in an industrial setting.
Specific precision has yet to be confirmed, but I'm working on getting data from an N3 sampling of known no-good bearings to compare to another N3 sampling of known new good bearings.
Please let me know what additional info you need.

Thank you.

If you're trying to measure the velocity of the outer race and can make contact, an encoder with a wheel might work. If you can't make contact, it might be possible to apply a magnetic field and then sense the changes in the field and see if they correspond to rotation.

But the elephant in the room question is what is there about the "off the shelf" instruments that drives the cost to 6 figures? Reducing the cost of something, even if it's sold in extremely low volume, by 3-4 orders of magnitude is often impossible.

It looks more like noise/vibration level measurement, nothing to do with rotational encoders.

@hites did you checked your PMs?