Locating the position of persons in room with arduino

Hello, I would like to locate the person ( margin of error 15 cm )of several individuals ( between 5 and 15 ) in a room ( 8m x 12m ).
And to know in real time the position of each person in the room.

The persons can be equipped with a wireless sensor or module to be able to detect their position.

Sensors can also be placed in specific areas of the room.

  • I had thought about ultrasonic transmitters and receivers, but there is a problem with obstacles, but also the impossibility to differentiate individuals.

  • The same problem with infrared.

  • I think by radiofrequency.
    Persons will be provided with transmitter modules and place in specific places receiver modules (or the opposite) and by triangulation detect their positions, each will have its own frequency and thus calculate the distances. But I don't know which Radiofrequency modules to use.

If you can help me you can tell me the different methods and sensors to use, starting of course with the less expensive ones for this project.

Thank you very much.

Posyx should work;

Posyx

The same question gets asked every few days, but in the past few years no-one seems able to produce a workable (and accurate,15cm type) solution.

If there were a cheaper solution (than Posyx) you would expect the project to be easy to find and ready made units to be cheap on eBay.

You could try image recognition on the Raspberry Pi. Adrian at pyimagesearch has done some object tracking tutorials which gives you an x-y position relative to the camera frame. You could use multiple cameras and triangulate the precise location of an object (the person's face). But 15cm resolution?

Some kind of hat with distinctive fiducials on top, with downward-looking cameras?

I would also suggest cameras as a reasonable solution. The room size is fairly small and you can outfit the room with sensors. So a couple of ceiling mounted cameras should be able to get the job done. Like SteveMann said it's something that would have to be done on something like a Raspberry Pi or maybe a PC.

found this two weeks ago

https://www.terabee.com/mitigate-covid-19-risks-improve-safety-efficiency/

Worth to email them, they might help you

In a 12x8m room every point is within 7.3 meters of the corners. At that distance a 15cm accuracy requires an angle precision of about 0.1°. With 1us timing, the 900 measurements needed for a 90° sweep would take less than a millisecond. In 4 milliseconds you could sweep an IR lighthouse from each of the four corners. Repeat the sweeps every 10 milliseconds and you can get 100 fixes per second per person. A BLDC motor spinning at 100 RPM could turn a cylinder with a mirror to do the sweep. The lighthouses could probably self-synchronize so no connection to the central server would be needed.

An IR strobe, sent from all four corners, provides a time reference. An IR sensor would detect when the four sweeps passed. They would send that data to a central server and some math would calculate where in the 12x8 room the sensor was. Each sensor only needs to see the strobe and two of the lighthouses to get a fix.