By and large I have seen very few impressive autonomous modes on any robot. Our team’s robot is definetly included within that description. Of the few that succesfully pick up more balls and place them into the high goal (so they get more than the 2 started with), the impressive aspect is the engineering that allows for this to occur by simply moving forward and backwards. That is, there is no apparent sophisticated programming or a large use of sensors. I believe I have seen only one team that tracked the white line, and that may have been dead reckoning.
What have you seen in autonomous mode and why do you believe it to be overall to be relatively simplistic?
My guess regarding the poor autonomous showing is that a large number of teams are rookies and have so many hurdles to overcome that autonomous programming gets put on the back burner, since you can’t really start programming until mechanical design is more or less complete.
In our region, we’ve seen 8 teams that have attended 2 scrimmages. As of the most recent scrimmage (2 weeks ago), only 2 had any autonomous programming at all. Some still haven’t finalized the chassis and wheels (they keep putting the robot together and taking it apart). It could be that there are some really advanced “sleeper” teams at our tournament in less than a week, but my guess is that on the average, teams that attend a scrimmage would tend to be somewhat more motivated than those that don’t.
We’ve didn’t really start programming in earnest until 3 weeks ago, partly because we were still working on major mechanical issues, but partly because we couldn’t get our software to work on any school computer, even having called Intelitek tech support 3 times (they tried their best, but the time zone difference was a problem) and getting the district administrators to work on unlocking the hard drive twice. We finally gave up and are using a borrowed private laptop.
Our programming is simple: dump 1 or 2 balls and push 4 balls into the low goal, all using dead reckoning. The other team we saw dumps 1-2 balls into the low goal. If we had more time, we might consider using sensors to navigate around the field to get more balls in. But one concern that we had was that too much running around the field might interfere with our alliance partner’s movements. As it is, we have to create a mirror program, in case our path on our preferred side interferes with our partner.
Our team uses dead reckoning also, if i understand the correct use of the word…
For the Florida regional, we could only score 2 balls into the high goal during autonomous…
but since we have had since december to get ready for Atlanta,
we have programmed it, by dead reckoning (correct placement of the bot and trial and error of how long to run motors/encoders?), to pick up 3-4 more balls and score them too in the high goal. It still needs some work but its coming along.
There was only 8 teams at the Florida Regional and all of them but one had some sort of autonomous mode. And the robot that didn’t seemed to be built right before the competition. lol. I would say that about 5 of the teams could score 1 or 2 balls into the high goals.
The ultimate autonomous would be that of legendary team 1114. I heard they could score their original 2 placed balls and pick up 6-8 more and score them in the high goals too.
Our team has a auto mode to get 6 in easily (first 2 + 1 pile), with a possibility of another pile as well, though it is much less reliable. We used dead-reckoning.
We do this by flopping our lid over the pile, with 2 already in it, and then turning 90 degrees right* and running them into the goal. The second half then turns left* and flops the lid over the second pile, turns 180 degrees and runs them into the goal. We have an option to not do the second half, and an option to be on a different side.
Well as the old expression goes, “If it weren’t for the last minute, nothing would get done!” We’ve finally started testing our auto codes for the new generation bot for our team (SPAM 3228). We have 8 different modes selected with 3 jumper cables. We still need some tweaking (of course), but our most robust scoring mode has the “potential” of getting 10 balls in the high goal if all the stars , planets, and moons align. We tested it about 10 times and got a low score of 2 balls (the 2 preloaded) and a high score of 6 balls. The mode was probably 3 or 4 balls.
We unstow, score the 2 preloaded balls (a bird in the hand is better than 2 in the bush), and then drive through the 2 closest piles picking up whatever we can get. We drive back to the goal and score what we picked up. If the robot gets knocked significantly off line when picking up a ball, we don’t score any more. We may run out of time and miss scoring some of the balls we picked up.
We may change the program to only go after the 1 pile close to the goal. Maybe being less greedy will actual allow us to get 4, 5, or 6 more consistently.
