I am looking into developing a small bot that could follow me around at a set distance (approx 3m) behind me. Ideally I would like to find a cheap pair of radio emitters and receivers that I could use to measure distance (2 small emitters placed on me and the receivers on the bot) This way I could separate the receiverson the bot to triangulate the position of me relative to the bot, this being used to guide the bot in my direction. Are such things available as I can’t seem to find anything anywhere.
Has anybody done anything similar to this who could share some of their ideas or findings. I have an alternative idea as follows but it’s not as solid. Comments on potential issues would be greatly appreciated.
Use an ultrasonic range finder mounted on a servo or stepper that scans left to right until it finds a target (me) and returns the distance and angle so as to guide the robot. It is then continually scanning left and right at the target until it passes the target and then reverses direction and so on – this would give an angle range of the target with the direction being between the 2 extremes.
Radio emitter/receiver are Radio emitter/receiver are only able to communicate, they don’t know their position… For that you’d need a gps module (or in your case, several, but you need really accurate ones if you want two on you… and i’m not sure you can find that)
Many thanks. But is there not a way you could send and receive a signal and measure the time lapse to evaluate the distance ? Or would this need real specialist equipment ?
I can see how your (second) idea might work, but I see many obstacles. The main problem is that you (your legs presumably) are not big flat square hard surfaces. And your environment (unless your are playing around in an empty parking lot) probably is full of such objects.
That’s why the ultrasnound distance ranger is so effective as an obstacle avoiding sensor in household environments.
I like your first idea better. It has been discussed here before. People just want to build robot puppies, I guess. Forget about accurate distance measuring. Just try to get your bot to see which direction to find the transmitter (in your shoe?). And maybe you can have the receiver measure the signal strength. Make it stop when the signal gets too strong. Say at 3 meters distance.
I don’t think anything like that has ever been tried here on LMR, so i really don’t know the answer. But personally i don’t think you can do that with the equipment we use, you know…the difference in reception is going to be very very very small…
You could try to see if google has something about specialist equipment, or else you could go for the gps, which sounds much easier and more effective i’d say.
That’s not exactly correct I’m afraid; there are several methods for determining direction and/or position using various radio transmitter/receiver combinations. The two main ones that come to mind are: • Scanning and triangulating using directional antennae (1 transmitter and 2 receivers, or 2 transmitters and 1 receiver) • Triangulating by measuring varied signal strength (1 transmitter and 2 receivers, unless you use mutiple frequencies)
Unfortunately the second option is almost totally impractical at such small scales… and to be honest I don’t see the first option as being too great either. In order to get any kind of usable accuracy from the sensors you’d need to have the 2 transmitters or the 2 receivers a decent distance apart, and making directional antenna at a small scale would be a right pain as well.
Let’s do some rough calculations to check the viability of this concept: • Speed of radio waves: C = 300,000,000 m/s (give or take) • Optimal distance between transmitter and receiver: l = 3m (from target to robot) • Minimum time between sensor readings: 0.2μs (this is one instruction cycle for a PIC using a 20MHz clock)
The time it takes for the signal to travel from the transmitter to the receiver is 3m/(300,000,000m/s)=0.01μs. In other words, even quite a fast processor with flawless sensors and programming would still be 20 times too slow to catch the delay from transmitting to receiving. This idea works for GPS because the satellites are very far away from the receiver and each other, but it just doesn’t scale well to the kind of dimensions humans are used to.
Thanks all for the comments… Much appreciated. ! I like the idea of ultrasound, i will research this a little and see what comes up. Again many thanks!
With a single transmitter and a sweeping, highly directional antenna it might be possible to pull this off. The angle of the servo when the signal reaches maximum strength would give you the relative angle from bot to target, and thanks to the inverse square law you can use the reception strength to estimate the distance.
about the second idea… i about the second idea… i have an idea into mind but i don’t know it it can be made. Something like: placing a tube in front of the receiver which grounds all incoming transmissions, basically this would let the sensor receive data only if directly facing the transmitter. Could this be done? I am afraid that data might be grounded even if they are directly facing each other (tx and receiver).
Now we’re getting somewhere Now we’re getting somewhere (well you lot are), i like the idea of this - any thoughts of where it would be possible to find a directional antenna ? Cheers.
Hmm, the problem with your Hmm, the problem with your typical reflective ultrasound is that there’s no real way to tell which object is your intended target. Has anyone ever tried using a discrete transmitter and receiver pair? The target could have an ultrasonic transmitter in their pocket, and the bot would then ‘home in’ on the broadcast signal using a sweeping ultrasonic scanner, or a pair/cluster of receivers. You could even encode a password into the broadcast signal as some sort of basic ID check. Theoretically you’d also get twice the maximum distance from the usual hardware, since the ultrasonic pulse only has to travel one way, not to the target and back again.
You might find it’s easier You might find it’s easier to just make one from some schematics on the net. They use them in ‘Amateur Radio Direction Finding’ events and for tracking tagged wildlife etc.
But maybe you could use a But maybe you could use a simpler solution : you carry several IR leds as a beacon and your bot has two IR receivers to seek it, and you use the ultrasonic rangefinder just to be sure your bot stay at a safe distance…
