IMU for tracking / FOG / single reference point

Good Evening-

First time poster- although I've used the arduino in other products.

I'm having a bit of a problem trying to figure out how to design a 'tracker' that would log a person's movement around a room. The person is carrying a probe that reads data continuously; I need to co-locate those readings. The tracker will be on the person/probe that is moving. Rooms could be about 150mx150m max.

The first approach would be to use the inexpensive MPU6050. However all the discussions I've seen on it state it'll drift meters within seconds, which would defeat the purpose. (We use them on Robots for FIRST but they aren't left to run free that long).

Then I started wondering if knowing a single, fixed point (such as a wide-area emitting LED) whether or not I could limit the drift by constantly updating that 'fixed' point's orientation- a 'star tracker' so to speak.

Lastly, Fibre Optic Gyros- pricey but if they fit the need and aren't ITAR controlled, they'd be worth the cost (I only need 1 or 2 units built).

And why the trouble? The areas I have to work in are radio-silent. No wifi, bt, GPS. I'm stretching it even by creating a 'beacon' for absolute positional, but I think I can get away with it.

I'm also looking at LiDAR as well to 'map' the absolute wall positions- although that would be a heck of a lot more post processing to figure out where and what was occluded at the time of capture.

Right now I only have the 6050 in hand; if anyone would have a suggestion or perhaps a better set of search terms that might pull up someone doing a similar effort that I could leap from, I'd be most appreciative.


How about a couple laser range finders? Have two 90 degrees from each other, panning back & forth a little to find shortest distance to nearest wall, give you a more or less real time x/y distance read out from wherever you call the 0/0 point.
Have them on a holder that keeps them pretty level so up/down angle does not add to the distance. Or use a third that looks up, panning in a small circle and autotilting as needed to keep the left/right platform level. (or a 2nd pair, panning 90 degrees apart, might be easier to code.)

Excellent idea!

We already use one laser range finder- but turning it sideways was always a pain to get map coordinates. That would certainly solve my data export issues- and with a bunch of them I could figure out coordinates easy enough.

I'll have to give that some thought too. More manual work but that would get it pretty close in the mean time and give me a year or so to get another solution up and running.

I suppose I should say imagine wandering through a full room of cubes and TV stands and whatnot with people moving around- that's what I'm trying to 'map' while walking around taking readings. Anything involving reference points would be useful but very difficult to always be able to 'see'.

Dangit, now I remember a project that used 3 point LEDs at different frequencies to lock orientation locations. That could work well.

Need to elevate your tracking device above the other stuff so it has clear view of the walls.
Maybe put a compass in it and the small roundtable so it can orient itself more or less to north as you move around and keep the x/w lasers pointing at the same wall.