Don't bicycle races have inspections? Safety or otherwise? I'm more familiar with motorcycle and car races where there are long lists of "class", safety, and weight requirements.
Would be trivial to detect either by weight or looking down seat tube.
Messing with the height of someone's seat by pulling it out before a race would ... not be seen very well. A lot of people can tell when it's off by a few millimeters.
Bike seat height adjustments are very simple: the seat is mounted on a tube, which fits within the slightly larger seat tube and then is clamped.
Getting the exact same position is easy: just use a Sharpie marker on the inner tube to show where it fits in the outer tube. Draw a second, vertical line across both tubes to make sure the orientation is the same.
Yeah, but if you do that before a race... someone's going to do it quickly and not do a great job. Like I said, it's not that difficult to detect a difference of a few millimeters, and it's annoying.
Weight is meaningless (for testing). The UCI has a minimum weight that is well in excess of what modern carbon fiber layups allow. So, all pro bikes have extra weight built in, usually on the order of several pounds.
But, pulling the seatpost or crankset/bottom bracket should make it pretty obvious if a bike has a motor hidden inside the frame.
And it appears the UCI has started using some sort of x-ray device for quick tests in the paddock area.
Various theoretical methods include moment of inertia in various axis and simple center of gravity measurements.
Hmm... moment of inertia and CG measurements indicate that dude's crank is a couple kilos heavier than everyone else crank.
If carbon fiber is RF transparent enough then low level microwave radiation tends to react interestingly with semiconductor junctions.
After EMP sensitive gear (if any) were removed from the bike, old fashioned microwave RF zorching or fancier "real EMP" could make the issue moot. Once the electronics and wiring are blown, a motor is just an expensive piece of ballast.
The last technological way I can think of is a careful magnetic analysis of the bike.
I would imagine a very permanent long term solution would be a careful stress analysis and small holes drilled into all frames for a COTS borehole camera.
Would be trivial to detect either by weight or looking down seat tube.