I'm conducting a DIY hobby/project right now, that involves a custom directional anemometer I designed and 3d printed. I've currently hooked it up to an old DC motor (with minimal cogging effects) I had laying around, and attached the leads to a V/Ohm meter.
I'm experimenting with placement. So far, the front left corner of the roof line appears to be optimal, though I may play around with placement towards the middle and rear. I'm currently using velcro, but intend to move to a magnet base so I can relocate it anywhere on the roof line. My current design is closest to a turbo type impeller that should help prevent side-draft deviation. It may not be the most efficient, but I think it will suit my needs for now. It currently kicks in at about 35mph at 1v, and at 65mph, I average 12v. This works well as drafting under 40mph is quite negligible on economy anyway.
I've already made a discovery doing this in only two runs. I'm getting pretty steady results without cars around, but while drafting behind big rigs, I've noticed that any adjacent vehicles basically kills about 1/2 the drafting effectiveness. On a calm day at 65mph behind a big rig, I get roughly 4-6v. That varies by half the voltage without cars. As soon as a car or truck comes beside me, it jumps from there to 7-8v.
I've yet to equate it to a linear scale just yet, but assuming 35mph is 1v, and 65mph is 12v, 5v should be close to 50mph. Next run I'll test out my theory and it will help me determine relatively accurate voltage to wind speed ratio. If it's higher or lower I'll have a curve to base it on. I'm also looking into adjustable analog meter that I can replace the digital voltage meter with. if I can find one that has a numeric scale of 0-100 that operates between 0-20v, that would likely be ideal, and can calibrate it to actual wind speed if the scale turns out to be a linear one.
I've considered going Arduino to do a digital calculation and logging, but I'm intending on keeping this project as simple as possible. Though, with Arduino I could use a much simpler anemometer based on magnetic switch or (hall effect) to make it more weatherproof, lighter and more accurate at low speeds. As soon as I start getting that complex, I start to considering using a canned Bluetooth anemometer. Those seem to start at around $70 on Amazon, and so far this project only cost me a few cents in 3D plastic filament.
I've also started 3D printing vortex generators. Once I get a handle on my anemometer setup, I'll start implementing those and seeing just how well they actually work. I have a theory that they can be useful in the front of the car just as much in the back to reduce drag.