PDA

View Full Version : Do you think it was good


Mechanical
03-31-2009, 10:05 AM
I thought this year was a bad year to introduce the system. On our robot you could do everything with last years system. Though i do see how it will be benifical in furture games.:]

JesseK
03-31-2009, 11:05 AM
I think it was a great year to introduce the system. The floor let all of us get away with so many low-quality nuance designs, it took the pressure off of most teams so they could focus on software. On top of that, some teams simply built cookie cutter drive trains, while many teams experimented with drive trains that have never been done before in FRC (330 Beachbots, for example, and 1885 though ours is a reinvention of a previously-seen drive train). The drive trains could mechanically have been done before, but much of the processing to control it may or may not have been eaten up by the old IFI CPU, which would have significantly hindered the amount of PID processing available for some of the feedback systems seen this year. Many teams were also finally less inhibited to try the wacky, crazy, or difficult code-wise. Such examples are camera code processing at 15fps, and many drivers stations that utilize the data received back from the cRIO.

RyanCahoon
03-31-2009, 11:46 AM
I thought this year was a bad year to introduce the system. On our robot you could do everything with last years system. Though i do see how it will be benifical in furture games.:]

I have two responses to this:

1. As has been discussed before, I believe teams not "fully utilizing" the control system is more a factor of the teams' unfamiliarity. I saw the teams doing some amazing things. The camera stuff couldn't have been done with the old control system, for example. Maybe issue a challenge to your programming team to step up their game :rolleyes:

2. I think the best year to introduce a system is one where the transition will be least felt. Can you imagine how many teams would be left out in the cold if this year's game required teams to get the camera working? My opinion is ramping into the new system (traction control? :p) is actually a better method.

--Ryan

AdamHeard
03-31-2009, 12:17 PM
I believe that whether or not teams utilized it's new features was much less important than whether or not FIRST had it adequately tested before season.

Nate Smith
03-31-2009, 04:25 PM
...whether or not FIRST had it adequately tested before season.
Testing is something that was definitely done..."adequately" might be able to be argued, but...I propose the following(speaking partially from experience here)...

Extensive testing was performed prior to even the release to the beta teams (example: I was at FIRST HQ at the end of September, and was able to see a working bot on the new system, as well as pick up their "testing" WLAN on a laptop.
The beta teams helped to refine the code and test for functionality on actual "competition" machines
The unknown came in when introducing the FRP flooring - the resulting static charges generated could not be anticipated by testing,
The driver stations were also left in a "set up and go" state during testing - the frequent disconnects and resulting ethernet port failure was not experienced under this environment.
The relatively last-minute change to include the WPA encryption caused problems in getting everyone up to speed as well.

I think as an interesting comparison, I would like to have a full 6-bot field of machines playing this game, but running on the IFI controllers, without any of the static protection measures implemented later in the season, and see how well they fared.

AdamHeard
03-31-2009, 04:53 PM
Testing is something that was definitely done..."adequately" might be able to be argued, but...I propose the following(speaking partially from experience here)...

Extensive testing was performed prior to even the release to the beta teams (example: I was at FIRST HQ at the end of September, and was able to see a working bot on the new system, as well as pick up their "testing" WLAN on a laptop.
The beta teams helped to refine the code and test for functionality on actual "competition" machines
The unknown came in when introducing the FRP flooring - the resulting static charges generated could not be anticipated by testing,
The driver stations were also left in a "set up and go" state during testing - the frequent disconnects and resulting ethernet port failure was not experienced under this environment.
The relatively last-minute change to include the WPA encryption caused problems in getting everyone up to speed as well.

I think as an interesting comparison, I would like to have a full 6-bot field of machines playing this game, but running on the IFI controllers, without any of the static protection measures implemented later in the season, and see how well they fared.

I may be hard on FIRST, but when it comes to the driver station, it was not tested adequately. I also don't believe that FIRST should be given a blank check in all areas that is, "Well they tried.... so we can't blame them". Merely leaving a wallwart powersupply plugged into the DS KILLED a field computer at LA. This is not heresay, this is not rumor, this was my team leaving a powersupply plugged into the DS, but not a wall outlet obviously (something we did as standard practice on the OI of IFI days), that caused an hour delay at LA. I had several volunteers inform me many times that WE had broken it.

The static issue is another (and one that can't simply be blamed on the field), we had issues with unreliable and flaky ports on the DS in our shop. This was when the entire system was on one board of plywood, nowhere near a robot or FRP.

There were also many random faults and failures that my team has documented; we'll add to that list at champs (Well, hopefully not) and then send it to whoever needs to see it.

I agree the static may cause some issues with the IFI system, but many of the irritating faults we've experience this year we're unrelated to the field.