Get distance from point? Ideas how?

I am going to build a quadcopter and I wonder if there is some way to detect distance to a source of some sort or get long-distance positional tracking (inside house). I was thinking that I would get a speaker that constantly send out really loud pulses of sound in a frequency not hearable for humans. I could then get the approximately distance to the speaker? Am I right? If I tested the device in a room and got the value, let say 5. And then I came back the next day. It would still be approximately 5 right.. If I use this and a barometer, I could get the x,y,z?

Is there any better way?

Thanks in advance.. :slight_smile:

You can get positional data inside from ultrasonic sound but not with measuring the volume of a loud signal but by installing 4 ultrasonic receivers (microphones) and put an ultrasonic transmitter on the copter. By comparing the timings of the pulse receptions you can triangulate the position of the quadrocopter. Theoretically you can do the same with 4 transmitters and one receiver on the copter but then you have to use different ultrasonic frequencies and you have to prefilter it on the receiver to distinguish the different signals.

What do you mean by "long-distance" when you're talking about in-house signals?

Triangulating position based on sound waves might work outdoors but not indoors. Too many echoes. Even outdoors you're going to get lost if the copter starts to yaw.

If you poke around on Youtube for examples of high precision positioning of quadcopters you'll see it's typically done with cameras mounted around the room. A computer then locates the copter in the camera picture and sends commands to the copter to adjust its position. As far as computer vision goes it's pretty simple to locate a copter if it's the only thing in the room with a bright (or specific color) light on it. The 3D locating using multiple cameras would be a bit of math though.

Triangulating position based on sound waves might work outdoors but not indoors. Too many echoes. Even outdoors you're going to get lost if the copter starts to yaw.

This is not correct, the students of the ZHAW in Winterthur, Switzerland demonstrated the contrary. PADA - Archiv
The indoor localization with an accuracy of 10cm is possible although not easily achievable.

Why would´nt my idea work (If I use barometer and get distance, and in some way the angle, I can get x,y,z)? I can´t use Triangulation because I will be in more than one room at once.. Does it work anyway?

Why would´nt my idea work

First: If you have a very good barometer you get about 1m accuracy. Is that enough for project?
Second: "and in some way the angle", there is no "some way".
Third: If you get the distance by measuring the loudness of a tone you might get an accuracy of about 20m in best case. Is that really acceptable for your case?

Using a barometer under a quadcopter is not the best idea. There are huge turbulences everywhere...