Quote:
Originally Posted by Greg McKaskle
Concerning reuse, the cRIO connectors are quite different from the friction fit of the previous system. The chassis is held by at most two fasteners, and the modules are clipped in. Disconnecting the chassis and leaving the modules attached to the robot is very quick. The next layer of connects, on the modules, are mostly D-sub based, and also quick to reconnect. Over the coming months, I assume many teams will come up with strategies for quickly putting the computing "engine" back into the robot with little to no rewiring. A backup of code images to install over enet is another important piece needed to put a robot into a different robot.
Greg McKaskle
|
First, can we safely assume that teams will receive additional I/O modules in the KOP each year? Or will we need to purchase additional modules just to keep robots in a relatively usable state?
I don't think anyone is questioning whether the cRio can easily be swapped into and out of robots. The issue being stated is that with only one controller the scenario you describe here won't work if our current competition robot (with controller) is in a distant corner of the country waiting for a competition. I can't hold a demonstration when the controller is one thousand miles away.
As for the mentors that mentioned pulling old controllers, our old controllers are already installed in older robots. Yes, we can pull one, but if a former-student-turned-engineer showed up thinking about mentoring, how are you going to explain that the robot they helped design, build, drive, and sweat over is an unusable pile of parts because you needed the controller for different robot. It's been hard enough explaining why we had to dismantle several of our early robots because the former school we worked out of wouldn't allow us to store them any longer.
While this may not be a big issue, one that will be is compatibility. Several documents are showing an ethernet camera with the new system. Anyone want to field an easy (simple and low cost) way utilize this feature with an older controller? What other future FIRST developments will cause further compatibility problems when utilizing older controllers with new robots? Do you also want to waste time rebuilding an older OI to use with the new robot?
How many mentors who mentioned using older controllers, have actually taken one of much older basic programmed RC's and put them into a new robot successfully? I know we have those controllers in storage, but they predate me and most of the current mentors on the team. I'm not even sure if we could find the basic IDE and compilers now. Could we get it to work? Yes, if we can find everything, but why should we waste vital mentor man-hours when the robot was functioning just fine with the original controller.
A two tier registration system is acceptable, but once again it opens the divide between the "have" and "have not" teams. The "have" teams can and will buy additional controllers if given the opportunity. Also, what happens if purchasing additional cRio controllers is prohibited or more expensive than current registration fees? Will "have" teams register a new "rookie" team and disband it after one season just to harvest a low cost controller?
I applaud FIRST for trying to find cost savings, but is this really a cost savings to teams?