2

I'm working on a project that needs a system that can determine approximate angle of an object related to it. I have anchor position (multiple anchors, but fairly close ~ 1m of each other). And object that moves in front of anchors (1m to 5m distance). I need to find approximate direction and distance to the object.

My anchor position is a platform with diameter of ~75cm. So anything mounted on there has a known position relative to other points

Here are couple of things that I researched that might work:

  1. Computer vision - this is probably one of the most straight forward solutions, but in my case it needs to work even with some minor obstructions - like putting the object in the pocket so it's not visible.
  2. GPS - I need to work with fairly close distances and it needs to work indoors, so this is a pass.
  3. "Bluetooth Angle of Arrival" - judging by some demos I have seen this looks perfect, but it is really hard to find anything on the market - it all is just inquiries for the product and it looks like it might be expensive.
  4. "UWB(Ultra Wideband)" - basically localised GPS system - so far this seems the best solution I have found, seems to be accurate and within reasonable price range.
  5. "RF array" - i'm seeing here and there that some people managed to achieve this with measuring signal strength with some RF array, but information seems lacking or I just don't know the correct terms to search for.

The main thing that needs to be noted that I don't need precise data, but consistent. I just need a general direction "far left/left/straight/right/far right" and for distance "close/mid/far".

Any other suggestions and/or weird ideas you could think of?

enter image description here

DKNguyen
  • 4,625
  • 1
  • 9
  • 22
somerandomusername
  • 141
  • 4
  • Can you provide a diagram? Are the relative positions of the anchors to each other known? Can anything be mounted to the object? At such close ranges trilateration using IR/RF-sound TOF difference tends to be simplest and most accurate. – DKNguyen Jun 02 '22 at 21:36
  • A diagram would be helpful, but also angles have three refrences, an apex and two "legs". On the weird side of things, how about laser devices? Without a diagram, that's just a wild guess. – fred_dot_u Jun 02 '22 at 22:07
  • A baseline, tape and protractor. – Solar Mike Jun 03 '22 at 07:36
  • @DKNguyen yes my anchor position is a platform with diameter of ~75cm. So anything mounted on there has a known position relative to other points. IR wouldn't work for me as it needs line of sight. But RF would be great, I would need to find a resource that explains how to go about it. – somerandomusername Jun 03 '22 at 08:27
  • @fred_dot_u I'm pretty sure you can get direction with just two anchor points. – somerandomusername Jun 03 '22 at 08:27

1 Answers1

2

You can use trilateration where the object is a transmitter that transmits an RF signal and ultrasonic pulse simultaneously and three receivers of known location measure the difference in time of arrival to determine the distance from the beacon. Then with three or four beacons (since the spherical shell representing distance from each beacon won't perfectly intersect at just a single point) you can determine the position in 3D space.

Normally in this case, sound travels so slow compared to RF that you can just treat the RF signal as being instantaneous but if you have latency from using a fancy RF signal that is comparable to the time of flight of the sound over the distances involved you will need to account for that. You probably don't want a fancy RF digital/serial RF transmitter here.

DKNguyen
  • 4,625
  • 1
  • 9
  • 22
  • Pretty nice suggestion would love to test it out. But I think it wouldn't work in my scenario. The space is changing and pretty cluttered, sonic pulses would echo in unforeseen ways. And I would like to have multiple systems in the same location, so they would collude. – somerandomusername Jun 03 '22 at 14:07
  • Well I could solve collusion with using different frequencies. But not sure about the reflections. Might just need to play around with it to find out. – somerandomusername Jun 03 '22 at 14:24
  • Might not be as big an issue as you think but you can also just ignore reflections by spacing out the transmissions sufficiently and only accepting the first sound after the light pulse is received. – DKNguyen Jun 04 '22 at 17:41
  • I think I found a commercial product for this system, https://marvelmind.com/ This looks similar to what you described. – somerandomusername Jun 06 '22 at 07:37
  • Yes, like that. – DKNguyen Jun 06 '22 at 13:15