[Tottanka]

Originally Posted by Yipyapper

The Kinect might use a lot of resources, but if it's on the driver station and connected via USB, there should be no data sent to the robot save a value or two, which is negligible. If it's on the robot and the processing is done via Wi-Fi, then you have a significant problem with bandwidth (most likely).

We originally planned on Doing Retroflective recognition with Kinect + Resbery Pie, but when we saw all the problems we switched to the Laptop's Webcam and driver interaction.
We even turned the RPI off, so there would be no problems there.
I too, dont see how it should pass that much packages through the field, but do remember the FTA coming to talk to us about it. There were no problems, but i just wanted to give teams a "heads up", that if not implemented correctly this code might be problematic.
That said, it's not that hard to implement.

On another note, maybe the teams who have tried this method can do some kind of an 'help all teams' stand at the championship, where we could set together premade codes for C, Java and Labview, and just come to teams and help them get those 5 more points. Sounds easy enough, nost laptops already have webcams - so why not. It's kinda like waht 254 did to 971 at SVR, isnt it?
Ideas?