Local positioning System (LPS)
There have been other threads discussing LPS using time of flight. It is difficult for the DIY to get time measurements in the sub nanosecond range. Another method that could work is to measure angles. With 3 or more stationary (I will call them Light Houses (LH)) LH the position could be calculated. I have done some testing of LEDs like the one used in flashlights. Pulsing them at frequencies from 10 to 20 K Hz. I used the same kind of LED as a receiver and transmitter and back biased the receiver through a resistor. Looking at the voltage across the resistor it was easy to detect the light over 50 Ft with an oscilloscope. This distance could be increased with a better receiver. This light beam could be swept a crossed the control area by a small gear motor and a pot could be used to read it position. The position would be transmitted to the receiver either on a RF signal or on the light beam. The peak amplitude of the light beam could be used to know when to capture the angle information.
Sounds like fun.
Are you locating in two or three dimensions?
What are the dimensions of the search space?
What accuracy do you hope to attain?
Servos use a "small gear motor and a pot" and can directly go to an angle specified, but I don't think cheap ones would have enough resolution to be practical in this application. You might be able to tinker with your concept inexpensively using them, however, and possibly rule out the approach before too much more time and money were gone.
HTH
a7
You are on the right track. Most people call the reference points "beacons", and using fixed beacons is a popular method of indoor localization.
Pololu sells a simple example of a direction finding beacon transceiver that uses IR diodes and photodiodes, which may give you some ideas.
Your comment “Sounds like fun”. Sometimes my mind get fidgety and I need to give it busy work to keep it out of trouble. .
What got me started was a thread where the goal was to control a device that could cover a lawn I assumed a lawn mower. The area would be from 10 Ft to 200 Ft from the LHs and an accuracy of about 1 foot. I believe that would require an angle accuracy of .28°. If the LH were sweeping back and forth across 180° that would be about 1 part in 630. I think a pot readout should be able to supply that and an Arduino AtoD would have enough accuracy. I visualized a Arduino in each LH whose job would be to read the angular position and transmit it to the mobile devices and control the direction of the motor. For less than US $5. There are dual shaft gear motors. Put the pot on one shaft and either a mirror or the modulated LED on the other. Add LHs to cover around the back side of the house or a larger area. Unlimited number of mobile devices.
Thanks jremington
That might be used to keep multiple mobile devices from bumping to each other.
It is a formidable challenge to pinpoint the angle to a light source, outdoors and presumably in full sun, to that accuracy. You will very likely need detectors with focusing optics and color bandpass filters.
The light source should also be modulated at a well defined frequency, as done with infrared remote controls (typically 38 - 50 kHz).
So what kind of fun do you want to have? 
You are reinventing the servo, that's fun but can you do better than
and maybe with laser beams.
I think some time laying out a proposed system of beacons might show up some limitations. Maybe even enough to drop the idea like a hot potato.
Place the lawn mower where it is at the greatest net distance, measured I have no idea how, and see what the angular resolution would have to be.
I'd have fun just trying to get a few of those doing anything at all like what would obvsly be necessary for your numbers.
I might start with cheaper servos and a very smaller area.
Please report your progress!
a7
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.