This is a link to video of 575s 43 (sometimes 44) point auto routine. In this run, they actually scored only 37 because a cube bounced out of the fourth goal, but you’ll get the point. The team wanted a second run for the video, but we had a lot of stuff to do and just didn’t have time.
Scoring: Six goals, eight cubes (sometimes nine) in goals, bonus cube knocked off.
Interestingly, since the line following code accelerates the robot if the error on the line is small, the routine always takes a different amount of time, ranging from 56 to 64 seconds. I believe this example runs about 61 seconds. They’ve tried speeding it, but it loses reliability. Having seen dozens of robots running Elevation for a one-minute auto at Dallas, I’m thinking 43-ish is simply the most you can do in one minute. In Dallas, 575s 37 was the highest total run, with second place doing 30.
Wow, that’s something to shoot for. One of our goals is to actually try and implement a line follower for this year’s competition… there are so many lines… and so few ramps…
How did you allow for the different light levels on the field at different competions?
Do you have a pre-match initalisiation routine to determine the white tape - tile difference?
I want to get our team trying a line follower routine this season.
They have an initialization routine that runs at the start of each match – it sets the gray level. Occasionally, it will miss the line and the team can only hypothesize that it was a lighting problem. After doing a 37 at worlds during the preliminary rounds, during their one and only auto match on the arena field they missed the line on their first turn and that was the game. No complaining, though, 2921 had the same lighting conditions and they scored more.
if our line following was anything to go by, this must have taken a LONG time to programme