SmartDashboard and vision targeting on driver station laptop

Our team is new to using SmartDashboard, vision targeting and running vision targeting on our driver station laptop as opposed to the cRIO. We want to implement these three capabilities this year. We’ve looked and it seems using Vision Assistant is the best way to design and build our vision targeting but we do not know how to integrate the resulting code with the SmartDashboard, so we can then communicate the vision targeting results to the cRIO. It appears that to extend the SmartDashboard we need to develop code in Java but Vision Assistant does not have an export to Java code function (only LabView, C and .net). Does anyone have suggestions (or resources we can refer to) for the most straightforward way to make these play nice together or will we just be stuck taking the Vision Assistant exported code, translating it to Java and then using it to extend the SmartDashboard?

We have the same plans for this year and have looked through mountains of example code but are still struggling. Any help would be greatly appreciated! Thanks in advance

I’m one of the mentors for Team 1280 (the original poster of this thread). We’ll share anything we discover back on this post for others who are struggling with this. Any help/suggestions from other teams who have had any success with this would be greatly appreciated!

Our (Team 1280) 2012 season is now over. We tried to get Vision processing on our laptop but because we couldn’t get the data we wanted communicated back to our robot, we eventually decided to place our vision processing logic on our cRIO in our C++ code. This worked well for us.

After many robot crashes in competition, and in the pits tethered we eventually removed all of our SmartDashboard logic from our code. We did this after we captured an exception on the NetConsole. The crash looks like it originated in the Connection.cpp code provided by WPI, which is used by the SmartDashboard. Once we removed the SmartDashboard from our code, we had no further failures. It’s too bad because we really liked what we could display on the SmartDashboard, and how easy it was to use. We also crashed during the build period while debugging when we were sending too many Log() commands to the SmartDashboard. Hopefully next year the SmartDashboard will be a little more robust/bullet proof.