# robot interaction.

• posted

i have an idea for my robotic project, have one robot follow another robot. How can i get one Robot 1 to send a signal out, which robot 2 picks up and determines in which direction and and distance he is from robot 1?

luke

• posted

Robots could have a lightbulb on their heads, and light sensors.

Robots could have speakers or beepers on their heads and microphones.

Robots could have perfume on their heads and chemical sensors.

Robots could have fans on their heads and wind sensors.

• posted

are there any examples for implementing rf transmitters and recievers to PICs? How can you determine the direction and distance from the transmitter?

• posted

You're talking about RF Direction Finding. If you wish to attempt it, you could try to set up a circularly disposed antenna array (CDAA) on the robot. I don't believe this is a practical solution because it's HUGE. The biggest problem in trying to miniaturize this is measuring the time difference in signal arrival. See

Next consideration against RF is what happens when Robot 1 goes around a corner? It's still transmitting. Robot 2 still receives the signal. Does Robot 2 know that Robot 1 went around a corner? No. Will Robot 2 attempt to follow a direct line to Robot 1? Perhaps. If it does, it will most likely run into something unless programmed otherwise.

Similar problem with Line Of Sight (LOS) tracking. Robot 1 goes around a corner and Robot 2 can't "see" it.

---Keith Lehman

• posted

• posted

For the RF example, I'm just talking about a single frequency transmitter with an omnidirectional antenna on the robot. It is not computer controlled.

Then two receiver antennas on the other robots that are directional. Just use some grounded plates so the antennas can only see in one direction.

The two analog values will get you a very coarse direction and range.

I'd play with the AM intermediate frequency of 455 KHz since you can get filters for like \$3.00 and regular transistors will work for amps.

?Ref=81653&Row=376854&Site=US It's pretty old-school analog, but I think I'll play with one in a week or so and let you know how it goes.

We are just getting going on a flock of 20 or so cheap robots that we want to use for "flocking" experiments. I think we'll try the sound localization since they will all sound like baby checks peeping for their momma.

We chose this PCB and parts kit

WinAVR and the GCC C compiler, and it's working great after just a few hours of installing and playing around. I'm a UNIX/Linux guy, so I can't use those dern-fancy GUI compiler tools.

% make % make program

and it all works!

• posted

Yes, yes, yes.

This is close to the thing I was talking about.

Very nice. But of course, I will want to put a processor in there.

• posted

?Ref=81653&Row=376854&Site=US>

O.K. that's simpler solution than the cdaa constructed on the robot.

How well you make the reflectors on the receiving antennas will determine how coarse the direction and range readings will be. I'm assuming that the antennas (with reflectors) will be mounted on motor assemblies and that they'll be set up to lock on to the strongest signal.

---Keith Lehman

• posted

And how does robot one decide which way to go?

You might just join them with a rope or rod?

JC

• posted

Gee, those things ARE huge. There are direction-finding schemes that are much smaller, using, for example, the four small antennas at the corners of about a 1-foot square that you may have seen on the roof of a police car. IIRC, these are for detecting and giving the direction of Lojack transmissions. I recall seeing the design of this in a Radio Amateur's Handbook or similar ARRL publication. The four antennas are switched to the receiver one at a time in a rotating order at an audio frequency, using RF signal diodes for switches. The switched antenna acts like an antenna on a turntable, first approaching the source (the received frequency goes up slightly due to the doppler effect) then going away from the source (the received frequency drops likewise). An FM receiver will have an audio output at the antenna switching frequency whose phase, relative to the antenna switching signal, indicates the direction of the received signal.

It might instead follow a direct line to a reflection of Robot 1's RF transmission. For one robot to track another's position reliably, use a combination of approaches.

-----