Follow me

Wissam Ch.

Hi ,
I want to work on a project on which a wheeled small vehicle controlled by an SX/28 MC has to follow its owner where ever he goes. (ofcourse·excluding steps and other obstacles ).

My current concern is how to track the owner path, and follow it ?

First thing comes to my mind is to use two GPS receiver modules from parallax .
The first GPS to be installed in the owners handheld unit, reads the location info every second and transmits it to the other unit on the vehicle , the SX/28·then moves the vehicle·and continously monitor·its GPS coordinates against the received ones from the owner unit .
A buffer can be used on the vehicle to queue GPS coordinates to move to .

I would like to have some input on this approach or any other approach that can achieve the same .

Regards




·

Mike Green

GPS is normally not accurate enough for this task (only within a few feet).

What kind of range do you expect (between the vehicle and the owner)?

Do you expect any significant obstacles between the vehicle and the owner
(like walls, shrubs, trees, glass windows, etc.)?

Wissam Ch.

Hi Mike ,

for the range it should not be very accurate , 1 meter (3 feet aprox) will be ok.

Walls and other obstacles you mentioned could exist .

since·GPS is not accurate enough of this task , what other solutions could be used.

Regards.

Bean

GPS is useful to get you close to the owner. Then you could use IR to get a direction TO the owner, and ultrasonics to determine how far away the owner is.

I thought about doing a similar project (so much so that I registered the domain name "www.followmecart.com"), but paying customers come first, so it was put on the back burner.

Good luck,

Bean

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
www.iElectronicDesigns.com

·

Zoot

Check out this thread: http://forums.parallax.com/showthread.php?p=612399

Steve Norris used a thermopile array from Devantech to track the heat of living thing (the array gives you a rough direction towards a heat source that is "warmer than ambient").

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
When the going gets weird, the weird turn pro. -- HST

Wissam Ch.

Zoot ,

I checked the thread and the project looks interesting , but what happens if there were more than one person in the vicinity of the heat seaking robot ?

i think my vehicle will abandon me in such situations and follow the Hotter source· [noparse]:)[/noparse]

Zoot

Another trick that can work, but requires hardware on the "master" -- setup two IR detectors with a "vane" between them so that the 'bot has to be pointed right at the IR source for BOTH to be triggered, and a slight turn either way only allows one to be triggered. Then have the "master" where a small battery operated "badge" that sends out IR pulses. That would let the 'bot track the master, and regular object detection can decide if the 'bot is too close. The problem here is that you need to have your badge on you, and you need to either put it on top of a hat or something (so that it can beam in all directions) or decide if the 'bot can only track your front or back.

I would have a small SX driven badge that sends out IR "bits" in a short sequence, so that the 'bot could differentiate between true IR tracking signal and IR "noise" from other remotes, sunlight, etc.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
When the going gets weird, the weird turn pro. -- HST

Steel

You could do this with 1 GPS Module and an RF Link.

The User can have a handheld that contains a GPS module, RF Link, and a main processor.

*Roughly*...
The main processor reads the GPS, and stores the coordinates in a "Forward, Backward, Left, Right"-type command set (with timing intervals). (reads new GPS coordinate, compares it to the last coordinates, calculates difference and tells direction)

The Directions are then sent to the vehicle via an RF Link. A processor on the tractor then runs the tasks based on the forward, Backward, Left, Right commands.

There will have to be some tweaking on the timing, but this solution *may* have the vehicle run the same patterns that the GPS Stored.

But Please...If this is dangerous equipment, such as a tractor or lawnmower...Second think the project. There have been people seriously hurt by trying to automate equipment.

Polywogs

···· Would it operate in an open area or would it have to track through obstructions?

Wissam Ch.

I wonder how in some movies they acurately pin point an object by implanting a transmitting device in it , and using a receiver unit , that can tell from which direction the signal is coming and how far it is.

If this is real then it could be a solution.
What you say ?

Doug Hale

Many years ago, I worked on RADAR early warning recievers. The Recievers used a multi element antenna. This worked sort of like sterio-vision. The receiver generated information about a given RADAR pulse based on the signals recieved by each element of the antenna. The info was, Time of Arrival, frequency of arrival, amplitude of arrival, and angle of arrival. By using this info from two (or more) antenna arrays, individual sources can be identified and distance to source can be computed. With angle of arrival and distance to source, we know what the sources position is relative to the receiver antenna. If you know its absolute position, then you know the sources position. And if you know that position over time, you know its course and speed, even if the receiver antenna is moving - we factored that in already.


Now for the really interesting part. If the source actually is a RADAR site, it is doing the same thing with the reflected pulses - figuring out where you are. It has a slightly easier job because it know the time of transmit and the time of arrival for a reflected pulse. Give a fixed propagation speed that the pulse took a round trip, we can easily calculate distance to target. The source also knows the angle of transmit will also be the angle of arrival of the reflected pulse adjusted by the time of arrival (depending on the source sweep rate). The amplitude of the reflected pulse is an indication of the size of the target (what percentage of the pulse was reflected).

However, if the target does not "passively" reflect RADAR pulses, but "actively" "reflects" them, then the RADAR source can be missled as to the where the target really is. This is the therory of STEALTH.

So back to your question, yes it can be done - with a very talented receiver - called a radio direction finder plus a bit more math - and so can a bunch of other real interesting stuff.

Can the SX do it - probably - I did it with a ~4MIP processor in the early 80's, but I also had a hardware floating point unit - but an SX doing FP in software would probably be about as fast - and if not, use a couple of them.

Doug

Wissam Ch.

Thanks Doug for the info.
You know where we can find such a talented receiver "radio direction finder " .
any online electronics store ?

Doug Hale

Build it. No, I'm not kidding. It's not that hard.

The hard part is the directional antenna array, and a little research on the web will tell you every thing you need.

If you have two directional antennas at 90 degrees to each other and the RF source is within that arc, and with range, the difference in signal strength will tell you where within the arc the source is. So if you have 4 antennas, you can now cover a full 360 degrees.

Now you either have a receiver per antenna, or you multiplex the receiver to each antenna, measuring the signal strength at each antenna.

Another way is to use a single directional antenna and mechanically sweep it through 360 degrees looking for the angle of the strongest signal.
I'm sure you have seen the RADAR antennas sweeping around and around or back and forth. Or the spy with a little box going beep-beep-beep
swinging the little box back and forth.


Now to get the distance - get th baerring from two points and triangulate.
Or you can cheat a little if you know something about the source - like its amplitude.

If you know the amplitude at the source and the amplitude at the receiver, the distance between them is directly related
to the difference in amplitudes. (if you build the source transmitter, have it actually send the signal strength - then you can compensate for "week batteries".)


An other way this is done is to have two HIGHLY accurate syncronized clocks, one in the transmitter and one in the receiver. The transmitter just sends the time. The receiver uses the difference between it's clock and the received transmitter clock to calculate the distance - (this is sort of how GPS works)


OK, its not that easy(you would not have read this if the first thing I sayed was it was hard ... LOL) - but it would be a lot of fun to do it - and the pride of accomplishment - and the things you would learn along the way ...

The concepts are simple:
There is a constant autenuation rate - the signal lose is directly related to the distance.
There is a constant propogation rate - the time it takes is directly related to the distance.
Signals appear to travel in straight lines (unless they bounce).

Now you get the fun of doing the engineering.


If you want more info, just ask, I'll see if I can help

Doug

Mike Green

Do keep in mind that you can't just build a radar transmitter and use it. In most jurisdictions you need a license of some kind and the equipment has to be approved for use. The main exception is if you have an amateur radio license. There are some frequency bands where specific equipment is licensed for operation by non-licensed operators. Cell phones, WiFi/Bluetooth/etc., Citizen's Band are all uses that work this way. Also there are provisions for low power operation in certain bands. Most short range wireless stuff works this way like the units that Parallax sells. You're responsible for not interfering with other uses of these bands of frequencies. If you do interfere and someone complains, you have to stop ... period.

Post Edited (Mike Green) : 11/15/2007 7:46:24 PM GMT

Doug Hale

Thanks Mike, for making the very appropriate "disclaimers'.
I wasn't talking about using RADAR - just some of the techniques that are used with RADAR - which will work with, say, the ISM band where you have to be tolerant of others interfering with you.

Doug

Mike Green

Well, the ISM band is certainly an appropriate frequency for RADAR. After all, early RADARs used something on the order of 300MHz if I remember correctly and that was "way out there" at the time. Now we think nothing of 2400MHz for all kinds of consumer goods.

Doug Hale

Actually, the K-Band (micro-wave) is in the 20 to 40 Ghz range.

In the early days, Magnatrons could produce micro waves but modulation was out of the question (other than on and off).