Building a local RTLS

Hi, I would appreciate you guys helping me figure out how to solve or at least get started with this “real-time location system” problem. I am quite new to electronics but am a fast learner.

Imagine a small outside open area, like a field where physical obstruction is to a minimum. The size is about 200 meters squared. In this field are people running around, perhaps playing a sport like football, or racing bicycles etc. I want to record / display in real time the location of every person on the field, with accuracy down to a few inches. The people I intend to track are willing to wear a small device to transmit/receive signals, and I can place static transceivers around the field, say one in every corner.

At this moment I am not too concerned about the software needed to display the data - I am more interested in how this might work with affordable electronics, and unfortunately this area is not my expertise, hence me asking for advice.

From what I have read, trilateration (not triangulation) is what I need. I should use four transceivers on the corner of the fields to collect the “info” from the people. The people on the field will use a transmitter / transceiver / transponder to automatically emit their info, at as fast a rate as possible, maybe at least 30hz. This info will be a unique ID and the signal strength / channel power of the moving object, for which I can use trilateration to calculate the positions.

GPS will not be sufficient as a solution here to due to lack of accuracy, expense and high power consumption. As for bluetooth, I don’t think it will give sufficient range.

Does this sound valid? Can anyone suggest improvements or critique this? How will I sync the transmissions?

If by “affordable electronics” you mean less than, say, US$5000 for a complete radio-based system, it may not be possible. A number of companies offer relatively inexpensive passive indoor localization systems that use wifi hotspots and similar techniques (which would work outdoors too, given enough hotspots), but no one has to my knowledge reported accuracy to a few inches. However, Zebra Technologies claims to have an active solution that does what you want: http://www.zebra.com/us/en/solutions/lo … cking.html Talk to them and let us know what the system costs!

Video cameras placed at various positions and elevations can be used to solve this problem inexpensively and accurately (e.g. tracking people with differently colored hats) but may not be feasible for your application. However, you might take a look at the surprisingly cheap Pixy tracking camera about to come out: http://www.kickstarter.com/projects/254 … ion-sensor

Hi jremington, thanks for the reply. Yes I’m starting to believe this may not be possible with with RF based systems, and yes affordable as in less than 5k.

I’m reading that RSSI or even RCPI is just not accurate or reliable enough, which is a shame.

Thanks for the heads up on the pixy camera though, very interesting. Not sure if visual detection is as practical - I like the idea of radio transmissions emitting in a radial pattern, where as a camera system you have to make sure the angle is correct to “capture” the whole scene?

yes, real time GPS-like accuracy via simple/low cost RF devices - not possible.

Some reading: “trilateration”, Time Difference of Arrival