Turning back the clock 2012

With most season over, and Champs right around the corner, it seems like a good time for reflection. What would you re-do about your team, if you could do this season over? Overhaul your robot completely? Spend more time getting sponsors? More awards focus? Small tweaks here and there? Did you happen to do so well that you’d gladly compete an entire season with the same exact team, same exact methods?

I think that for our team, Team 2410, we vastly underestimated and pushed aside the bridge. Our manipulator, as seen at both regionals, was mediocre at best. If we would have realized the absolute huge advantage that CoOp gives you, we would have invested in a stinger type device.

As well, I would say the next time precision shooting comes around, we will vastly expedite our fabrication time, so that we can get in as much practice and tuning time in as possible.

Oh, and get a heck of a lot more time spent on our Chairmans video…

Your thoughts?

*Edit: Yes I know there are threads here and there about specific parts of your robot, but I want an entire team review. Anything about this season you would change.

3397 would have moved our ball tower to the front of the robot, because our guides to lead it to the center didn’t work well. If we could go back in time, we would have switched out a Victor that failed on us during competition, causing multiple fouls. (We couldn’t lift our bridge arm, and the piece that connected the two sides together fell off…yeah). Speeding up our ball lifter would have helped too. We used a windows motor so it was kind of slow.

I would have the kids test the Chairman’s DVD to make sure the sound burned properly, so we wouldn’t have a silent video playing as we walked down to the field.

Oh man, where to begin!

  • Not lock ourselves into a poor design in the effort of staying on track.
  • Not make a dual axle shooter. It doubles complexity, weight, and fixing time while making backspin harder to acheive
  • Never ever use Double Doozy Gearboxes for Banebots RS550’s. They removed the listing that they were compatible after we called Andy and described our eternal problems with them.
  • Figure out the underlying problem with our shooter rather than trying to work around it (We are currently using 4 FP level motors in our shooter, when most teams have had success with much less power)
  • Test and figure out the ideal variables rather than using the “it works? Let’s use it!” philosophy that got us a shooter with a minimum shooting distance of 15 ft
  • Design a bridge manipulator that actually worked when we got to competition
  • Design a stinger or some other balance assist device to include with an over the bump collection mechanism.
  • I can’t believe we missed over the bump collectors but we did. Completely flew over our heads.
  • Actually checked in with the Chairmans crew before the evening it was due. The 1 day delay saved us.
  • Invest in autonomous over teleop. We ran the statistics, and at our first regional, if we had a bottom basket autonomous working every match (which required very little tuning of our shooter), we would have seeded 2nd rather than 6th.

At the same time, there were a few things that I wouldn’t change if I were to do it over. I would stay a wide robot. I would keep our excellent frame and drive system (pushiest treads in FIRST). I would keep the great increase in outreach and programs that won us the Engineering Inspiration Award. I would play in game the exact same way (Coop all the time) and I would pick the exact same alliances that we triple balanced with.

We tore off our launcher the second day after determining that it was more of a hindrance than a point scoring mechanism. Originally, the launcher was actually messing with our kiwi drive because the weight distribution took away traction from one wheel, disabling most omnidrive functions and essentially turning the drivetrain into tankdrive+.

With the launcher off, kiwi drive worked beautifully (this was the third time ever that it ran without top weight) and our bridge downer was functional as well. Attempts at the practice field proved we could actually get up the bridge with kiwi drive and balance within 30 seconds (this was without any formal practice). The goal then was to get over and try to be a defender bot, which I think may have been significant given most defender bots were running tank. We wouldn’t have as much pushing power, however we could still disrupt some shooting and run around the field quickly (the toggle-able 67% speed scalar [necessary to allow on the fly turning] I had up was still surprisingly swift imo). However, we had no practice with it, and the first time that we actually tried to get ourselves on the bridge (which was our last game :frowning: ) didn’t work out so well due to the viewing angle and our relatively short arm. I just had it, then messed up turning and fell off (again, no formal practice at this), then the team’s other driver told me to let 4232 get up. I honestly think that if we had the launcher off first day, we might have won a bunch of earlier matches due to the ability to balance and would have been a better candidate for quarters :o

Well we had no progress on the shooter during week 1 because of bad prototypes, and as a result, we did not have a working shooter until 2 days before SVR.

We would probably also have a much shorter robot and a wide based robot rather than a long robot.

Also, a slightly slower intake and robot may have been better, it is quite difficult to control balls that go from ground-shooter sub 1 second.

In addition, we probably would not outsource such a critical part to our new sponsor. Although the part was awesome and worked as it was designed, it had a huge turnaround time, preventing our robot from moving until week 6.

Finally, I probably would not delegate a task of such great importance to a person who was unable to CAD until week 2.

The quarterfinals at Lenape… :frowning:

Lower center of gravity. Our robot was around 60 inches tall, but the motors for the shooter were only running at 40%. We could have easily mounted it much lower and simply increased the power, making it easier to balance.

Perhaps a different drivetrain as well. We always build a standard, 4-wheel KOP system. In the future, we need to consider things like 6/8-wheel drop-center or different wheels to aid in maneuverability (turning, specifically) on carpet.

We could have taught some kids how to program before the beginning of build season, so that all of us (myself included) weren’t still learning how to use LabVIEW 4 weeks into build season.

More time for driving practice would’ve been nice as well. We always end up bagging the robot at midnight on the last day, with the motors still warm.

With more thinking, I’m sure there’s a not of non-robot things that could have been better, but I’ll save that for later.

Robot things I’d like to change would be making the whole thing shorter for sure. I get a little worried about it tipping sometimes, and it rocks a LOT when I go full forward to back because its CG is not very low. Not that I do that very often or need to, but it just limits how aggressively I can move that robot. It doesn’t like to stay flat on the ground if you do 360s either. It starts to jump. A nice low robot would have been nice.

The weakest part though has to be the transportation from the intake the the shooter. We had this complicated serpentine path with all these crazy ideas so that we could control the balls inside the robot well, but it ended up working nothing like that. Retrospectively, it could be hugely simplified, a lot lighter, easier to maintain, easier to control, and much nicer looking.

Structurally, I’m not thrilled about the compound angles on the tower’s supports, especially since we ended up mounting some things to them. We have 8 pieces of 1x1x.063 tubing comprising our tower, and that just seems like way too much. I’ve seen some great designs that keep the weight much lower, make a nice use of lexan instead of aluminum, and do some other cool things. Down low on the frame, it really isn’t all that strong. The front bent up a bit when we played some aggressive defense on the side of our robot that we take balls in from. Our bumpers mounts are made in a way that pretty much forces them to bend instead of transferring the forces back to the rest of the frame.

Our use of polycarb looks pretty sloppy. The polycarb that’s supposed to guide the ball is completely unsupported and deforms when a ball goes through. I was going to design some aluminum rails to supports it, but the CAD is a huge mess and I have no idea what shape the polycarb is supposed to be. We plan to invest a little bit more time post-Championship to redo the polycarb and the sponsor decals. I really wish we could have done it before Champs, but there simply isn’t enough time to tear out all the polycarb and install new plastic just for cosmetic purposes.

Our bridge lowering device is a little hard to use in terms of lining it up because it’s not that long (6" or so from the frame perimeter at most), but I guess that’s why we got away with using a window motor to lower the bridge. I doubt we’d have enough torque if we had a 14" appendage.

The shooter is pretty good, but I still can’t believe we put so much effort into using the Kinect on our robot to aim. It never did anything for us that the Axis camera didn’t, was more difficult to mount, forced us to put it off to one side of the wheels, and was about a full pound heavier than an axis camera. My paranoia as a driver about the weight way up to of the kinect, from what I knew about the software, and from what I’d heard other teams were doing, I didn’t want to use a Kinect starting from Week 2. But 10 weeks later we’re finally going to scrap the Kinect.

The drive trains are pretty solid. I wish the robot could be pushed across the floor by hand when it’s off, but that’s barely an issue.

Practice time would have been good, but it’s pretty hard with very limited shop access after build, and no real space for driving. We nabbed a carpet from SVR, but we don’t really have space to lay it out. Maybe we’ll steal the old gym next year for some practice.

I could probably spend hours coming up with things I wish we did differently, but I guess those are just things I’ll have to improve for next time (whatever project, organization, or job that may be).

Same. Our season…didn’t exactly go swimingly, so this is what we’ve been up to. Asking around and seeing what other teams have done well, not so well, and what they would improve upon. We know our mistakes, and we gave ourselves props where it’s due. However, we also want to learn from the best, avoid mistakes from the rest, and have one heck of a season next year. Only a little bit because it’ll be my senior season…:rolleyes:

-Drop down intake
-A little more time to program

-Make sure all wires are securely fastened (Montreal Regional, ughhhhh!)

We gave up on the securely fastening wires down and just hot glued them in…seems to have work out pretty well.

Pre-registered for Championship.

I’d have pushed for the CADder-in-chief to more closely follow the working prototypes demonstrated by the design subgroups, instead of freshly designing everything for the competition 'bot according to his own ideas.

Going along with that, I wish I were in a position where I could insist on a student being in charge of CAD.

If there was ever a year for it, I wish we had as well.

In terms of the robot, Fat Swan 2.0 would have the following changes:

  • Same overall shooter design (single axles, dual wheel, 2 position pnuematic hood)
  • Slightly shorter chassis, small robot, still “long” configuration ~30"x24"
  • Full bumper coverage, mounted as low as possible
  • Single speed drive, geared 8-10fps, 8wd, Colsons, brakes
  • Shorter overall height
  • Over bumper pneumatic collector/bridge manipulator with pneumatic “Stingers” (like a 228/971 combo)
  • CG balanced heavily opposite the collector
  • Shoot opposite from intake (one of our biggest mistakes)

Enforcing our weight safety factor much stricter

Avoid two side pickup, (took one off to help make weight)

Tank drive instead of mecanum (could actually drive up ramp fine, but could not help another robot up or “control” a multi-bot balance)

Move the build season one week forward (too much discussion)

Have mass to add a stinger once we saw 33 and 148 in action

Looking back I wish the team could have

Removed upper conveyer

move shooter to top of bot

made the frame smaller

In 14 years of competing we finally have a short list of “coulda, shoulda, woulda’s”
We have: low cg (34" high bot)
single wheel shooter/adjustable hood/rotating turret
shooter opposite intake
heavily weighted opposite intake (70%)
stinger for triple balance
a decent tipper --it works well, but it takes about 3-5 sec to get on a bridge. If we had our druthers, we’d prefer to be able to drop a better wedge and run at 3-5 fps right onto the bridge.

I wish I hadn’t been so willing to give up on what I knew would work, but hey, one person can’t (and perhaps shouldn’t) force everyone to do things his way.

I wish we had formed a subteam to build a bridge manipulator.

I wish we had finished our shooter earlier so we could have actually tested it to be able to improve compression, put on only one set of wheels, etc

I wish our collector fed directly into the shooter.

Non-essential but I wish we had designed a stinger, a drop down collector, and put less emphasis on being able to drive over the barrier, which we were not able to do anyway once everything was assembled.

I wish we had built real bridges to test with. rather than light plywood ones.

Additionally, we had considered a wider ball collector, but did not think we would be able to prevent jamming. I’m not really sure how teams were able to do this, could someone who did it well explain their design? Thanks.

This thread seems to be (as one would expect) dominated by teams who didn’t do very well and/or underperformed.

As we all know, everyone can improve and continuously get better. I wonder if any of the “Elites” would weigh in to this discussion on things they wish they had done better or focused on more.