The Access Points on the 2009 Control System

[gunsanbob]

to tell the truth I would rather watch all the robots

[neutrino15]

I think that if anything, they would stream the robot feeds to their own media server, and project certain feeds onto the big screen.

More likely (hopefully), they will let us communicate with the field, as well as other robots on said field. Then you could wirelessly tell the field, for example, to rotate something, or open a door..

No matter what, though, we will be able to communicate robot-robot (this IS standard TCP/IP).. I say we develop a standard for inter-robot messaging.. Depending on next year's game, we would possibly be able to group up and accomplish certain things together.. Even if it's just to tell everyone on the field some tidbit of information.. Think about it:

Operator Interface -> Robot -> Field Network -> Any Other Robot -> Any Other Operator Interface!

Or

Robot -> Field Network -> Any Other Robot

[yongkimleng]

Hmm my take on this:

1. use smarter nodes for APs. Like those USD$30 fon/meraki devices.
Allows encryption, filtering, etc. We certainly do not want rogue data interrupting the matches.

2. UDP for control. All realtime control protocols use UDP.
(Whats UDP? Google )

3. QoS is necessary. Set control packets to highest priority, video stream to lowest, debug messaging medium, etc

4. Use various channels. There are approx 13 channels, of which some slightly interfere with each other. Now we approx have 54Mbps of capacity per channel. Designing the system to operate at a minimum of 11Mbps or lower would be good coz, in a case where a channel is saturated with colliding data from other networks on that channel, your system will likely to perform constantly. We can have 1 team using 1 channel per match, i.e. when 2 matches are going on simultaneously they would be using 8 channels.