multiple robonovas

[brindy]

has anyone tried to do anything with multiple robonovas?

specifically, i'm wondering what the best approach to having two or more robonovas communicating with each other might be, if it is feasible in the first place.

the kind of information i want to be able to communicate is their relative positions and orientations. assuming that i can get that information from an individual robonova, how would could this be communicated to other robonovas, if at all?

thanks in advance for your thoughts / suggestions

[Robo1]

With out a method of getting there position then you can transmit it. The simplist method would be a indoor GPS system using 4 sensors in the corner of the room and a receiver on the bot to work out it's position.

To transmit this would require a simple circuit with a chip that broadcasted it's location.

So the short answer:

It would be quit hard and expensive 200 pounds but once set up you could do some pritty cool things (it would be able to find the fridge and the beers easily).


Bren

[brindy]

With out a method of getting there position then you can transmit it. The simplist method would be a indoor GPS system using 4 sensors in the corner of the room and a receiver on the bot to work out it's position.

To transmit this would require a simple circuit with a chip that broadcasted it's location.

So the short answer:

It would be quit hard and expensive 200 pounds but once set up you could do some pritty cool things (it would be able to find the fridge and the beers easily).


Bren

thanks, bren.

i'm not too worried about getting the actual positions - there are various techniques to do that using sensors already available combined with some math - what i really need is for robonova a to tell robonova b where he is.

another option would be some kind of transponder which each bot could scan for.

any ideas?

in case you're wondering what i'm up to, i want to see is some autonomous robonova fighting and also try out other experiments such as flocking - though i doubt i'll have enough robonovas to give a practical demonstration of that!

[Robo1]

I'm looking in to autonomous robo1 fighting but I'm going to use a camera.

If you take a reference frame at the beginning of the bout you can use it as a reference to work out how fare away a bot is e.g. comparing pixel number. the more the pixel number the closer it is.

I've just got a camera CMUcam1 from http://www.seattlerobotics.com/ that I'm going to try. So if it works out well I will post some info on this site.

bren

[Pev]

Well you could just stick a tranceiver in each RN-1 on their ETX/ERX ports (maybe even a bluesmirf if the latency isn't an issue - which it probably won't be with Robobasic comms).

Then have RN-1(A) act as the master and establish a connection to RN-1(B). Once connected you could design your own "protocol" for the data transfer - like first byte is north/south position, second byte is east/west position.

Pev

[DirtyRoboto]

How would the RN know what direction it is pointing without an electronic compass?

If you did have an Ecompass installed then you could work out the distance by pinging the other bot and know the direction from the Ecompass.
This would be the case if as pev says, you have an RF tx/rx installed.

The coding for this using the limited Robobasic package would be a nightmare, if you added another bot then you might aswell book the next 2 months off of work (or school) and get coding.

I was looking at implimenting RFID on the RN1. The bot would know the distance to a known object and also execute the correct routine to interact with the object.
If I myself wore a RFID tag then the robot would sense my approach and act accordingly ect.
If 2 RN1's had RFID then they would be able to know the position in relation to their own of the other Bot.

They would know that the red ball is RFID 000476 and the blue cube is RFID 000250, that Bot A is RFID 000111 and Bot B is RFID 000112 and so on.

This is easy logic to program into RB and takes alot of the donkey work out if object interaction.

HTH.

Marcus.