Hi,

Just hoping to get some ideas and advice on a project i'm doing. I am looking to track the movements of a participant in a project and want to create a patch that will allow them to be tracked by connecting a ultrasonic range finder, compass sensor and accelerometer and mapping out the area they will be walking around. I was wondering the best way to go around this and what calculations would be useful for me to use? I am then going to be using this information into a visual mapping and create a gps like visualisation.

Thanks very much in advance!

I wonder if Wi-Fi could do it for you by triangulation..... probably not. I know that wifi drop and connect can be fine-tuned, but with a portable computer like an rpi?
Tracking has been attempted for moving lights. Originally the "actors" carried infra red transmitters, but of course there was quite a lot of ir noise on a live stage. How are they doing it now?... I don't know, but I think that they eventually found a solution, and there is rarely any point in re-inventing the wheel. You could try ultrasound maybe, with triangulation. I would think that triangulation by some means is the "solution". Calculating by vector (as you suggest) will only result in misery, as all and any errors will be compounded.
David

@JP848_48 If you've got Apple or Android mobile devices, try MobMuPlat http://www.mobmuplat.com/ You can use MobMuPLlat to access a phone or tablets' GPS and accelerometer info and connect to pd with it.

@rjpg I tried working with GPS before, but the precision was absolutely appalling... In my case, I was creating a sound installation with an area of around 10m x 10m, and I decided to add large fiducial markers to wireless headphones and have a wide angle webcam shooting from above.

Our current GPS is only accurate to 7 metres, which might be bigger than your "stage (donated by the U S of A in 1983..... In 1983, the GPS ceased being solely a military system and was made available for public use. President Ronald Reagan declassified the GPS system in response to a tragedy involving a civilian airplane. Korean Airline flight 007 got lost over Soviet territory and was shot down by Soviet fighters.
Read more: http://www.maps-gps-info.com/gps-history.html#ixzz3WktAUrIQ )....".
The newer, European, non-military system is supposed to be accurate to 3 meters, but their satellites keep on "mis"-launching.
3 Metres is good for "turn left at the traffic circle" but probably not good enough for your purpose.
Gilberto's idea is probably the easiest to implement, with an overhead camera and movement recognition. It is already implemented in Video Trigger (by Zach Poff ), but only with 6 sensitive areas. If you run it twice (on 2 computers) with 6 horizontal bands for one, and 6 vertical bands for the other, with messaging between them ( you can use OSC and Midi if you run it as an add-in to H.C.Gilie's VPT7_01) then you would have 36 recognisable positions for your "actor".... enough? More cameras would give you greater definition.
David.

@whale-av I have the impression that the accuracy of the GPS depends a lot on the place you are (I may be wrong here), and also that for small sound installations it's totally unusable. Here in Prague, Czech Republic, I could hardly get it accurate to 10m, so I ended up leaving this method completely behind.