Hi, i am from team 3098 and we are looking to improve our vision processing in autonomous. We have looked around and want to switch over to grip for our vision processing, leaving vision assistant.
We have done some research and know that grip has trouble communicateing with the feilds FMS system. We want to use a coprocessor to handle the vision but we do not know what to use.
We would like to talk to other teams that use grip with a coprocessor and would like to know what you use, how you do it , and how successful you were.
Team 294 very successfully used GRIP on a Kangaroo on the robot. The only real issue we ran into was forgetting to plug it in a couple of times!
Regarding GRIP communication on the field, issues teams ran into were due to NetworkTables only using mDNS to try to find the robot, which occasionally had issues on the field. The main workaround this season was to use a fixed IP for the robot, but there are other plans in the works for next year to help fix this, such as having NetworkTables attempt to connect to a list of server addresses in round-robin fashion (e.g. USB IP, fixed IP, mDNS, DNS address).