Looks AWESOME, Scott. MAR is going to be pretty great this time around. I saw the 5 disc auton a few days ago and knew you’d have a 7 disc in store. Fantastic stuff, really looking forward to seeing it compete at TCNJ!
Love it. For the 7 disc auto are you shooting 3, driving over the pyramid discs, backing up to load those 2 and the 2 center line discs, and then pulling up to shoot 4?
There are enough teams showing a 5 to 7 disc auto that I doubt any alliance wins a division at champs without one on their alliance.
I love this bot but I am so confused at how it will do a 7 disk auto with the shooter and the pickup on different sides. Any more details? If not i’m sure I will be amazed this weekend.
Edit: Does it drive over them to pick them up cause that would be epic.
Yup! We start at the back of the pyramid, shoot three disks, drive to the goals (over the four other disks), then drive backwards (collecting the four disks) and shoot!
Definitely a cool motion path for a seven disc. When you drive over the discs, do they contact anything under the bottom of the robot? Or do they more or less stay where they are?
They more or less stay where they are. The 5-disk mode is easy to line up perfectly every time (3 shots + 2 under pyramid). We’ve been eyeballing the alignment for the 7-disk without too much hassle. If one of our wheels hits the edge of the frisbee they “self-align” and center themselves between our wheels without disrupting our path significantly.
Antonio implemented the Killer Bees autonomous scripting code for LabVIEW (http://www.chiefdelphi.com/forums/showthread.php?t=95120), so a lot of credit goes to them for sharing their work. It’s been tremendously helpful, and really made the process of fine tuning & implementing multiple routines much easier.
Very nice design, although I’m terrified for your cylinders. Are they able to rotate about the lower point or are they fixed in place? If they are fixed I hope you either drive very slowly to grab the lower bar or that you bring a few spares.
Yes, the pistons can pivot backwards around the mount. The red elastic cord in the picture pulls it into the upright position when there are no opposing forces.
LED Ring lights from superbrightleds.com, same vendor as almost everyone used last year. We’re using a 60mm and an 80mm green ring. They’re epoxied onto the plastic camera housing. One is plenty, but we like having a bit of redundancy/overkill in that instance. We caught a stray frisbee on Saturday that killed 3 of the LED’s. We might try putting a clear plastic shield over the LED’s to prevent that in the future. Last year we left it exposed with no issues.
Yes, we’re using it in both modes. We use the data to calculate distance and angle from target. That data feeds into the shooter wheel speed control and a “turn to target” function. The turn to target still needs a little tuning for autonomous.
I can attest to the fact that these lights are more than enough to illuminate the targets and then some. For those of you that are unaware, 2590 was nice enough to give us their spare lighting setup after ours (60mm and 80mm standard KOP style ring lights) was destroyed. The lights were noticeably brighter (unbearable at 3’ or so, to be honest) and seem much more well made. From what I could tell, one of the ring lights has about the same output as our original setup, which makes us feel much more confident that we’ll be able to illuminate the target if we lose a ring of lights… again.
Also, for those of you with any doubts, this machine is the real deal. Dead accurate shooting, effective floor pickup and a smart, well prepared competition team too.