Team "Best Practices" in a Post-Bag World

#1

There is a nice discussion thread for proposed rules to implement the “post bag” environment. Regardless of how that all plays out, this change will affect how each team goes about designing and building their robot(s?). It is likely to affect the amount of time spent on game analysis and prototyping, meeting schedules, build season milestones, competition schedule, planning for change, etc. I have a few thoughts to share, but I would really like to learn hear what this community thinks.

As for myself and my team, I really don’t see us starting over from scratch to build a “me-too” robot mid season (please God, NO!). I fully expect that we will make more mid-season changes than we have in the past, though. A few of my thoughts regarding how to prepare for this in our robot design/build:

  • Continue to build a practice bot. The benefits still outweigh the costs. Compete with the “fresh” bot. Beat the living daylights out of the practice bot.
  • Modular Design. We already do that somewhat, but we will need to take it to a new level. It is hard enough to implement a new end effector (etc.), but can we replace the elevator/pivot arm that the end effector is mounted on? What could we change on our “baseline” elevator that would make it more adaptable to some as-yet unforeseen end effector?
  • Spend some extra time thinking about the drive chassis and try to get it right. I expect that scrapping the drive base and starting over mid-season is just about the last thing we would ever want to do. We are probably going to be stuck with our initial choice for length, width, ground clearance, drive type (tank, swerve, mecanum, etc.), frame openings, etc. I don’t have any great insights on how to pick the most adaptable design solution for the chassis.
  • Make it easy to add new stuff. Drill plenty of extra holes in structural parts, using a rational pattern. Install more motor controllers and pneumatic solenoids than you need for the baseline robot. Its a lot easier to do this stuff before the robot gets fully assembled.
  • Be paranoid about weight from day one. Try to leave 10 lbs for that cool flux capacitor you might see in week 3 or so.
24 Likes
#2

Great thoughts. I would only add the necessity to implement an iterative design/build/test/deploy process. In that process, I would keep Week 0 as a “shakedown” run of the robot, then iterate with each scheduled event as a formal robot release point.

#3

We’re going to dedicate a significant amount of the build season to CAD and prototyping, but we were going to do that with or without a bag.

Honestly I think the teams that handle the post-bag world the best will be the teams that don’t drastically change how they operate. Keep your timelines mostly intact, and use the “extra time” for driver practice and extra tuning.

The teams that let the work fill their available time are going to struggle.

12 Likes
#4

For veteran teams in 2020, I would suggest to be cautious and act as if there is a bag day for the sake of robot build schedule, etc. Obviously, after the six weeks you can practice and iterate. Then, you can evaluate any possible schedule changes for 2021 once you have a no-bag season to compare against.

For 2020 rookies, hopefully there are some mentors that have FRC experience and can help the team get where they need to be.

I agree that most teams won’t have major practice differences compared to past years.

4 Likes
#5

I totally agree that for teams which are already bagging a working or near-working robot and using the withholding allowance for parts developed/iterated/exercised after bag, the changes should be minor. The main thing is that you will be able to do a bit of streamlining as you leave the specific “bag work-around” activities behind. In some cases, that’s one less robot or one less drive train to build. For others, it means you’ll be able to iterate on one chassis while driving the other.

For teams which currently really “stop build” and don’t build/iterate/practice on their robot after bag, (or meet to build a pit and other non-robot items), there will be some more serious decisions to make. Build the pit before kickoff and extend build season, or carry on as before? Spread the work out, press with the extra weeks, or limit yourself to the old 6-1/2 week build? Or somewhere in between?

1 Like
#6

Implement the six week build season with your own team. Your goal is to have a finished, working robot for week 0.
Go compete at a week 0 event. Use the extra time for testing and practice.
Don’t use the extra time for a longer build.

7 Likes
#7

Confessions of an Average Joe

We are a recovering build-centric team. Build gets the lion’s share of our collective energy, sometimes to the detriment of other team functions. For several years we’ve been working to fix our problem, so that outreach, programming, drive team training/practice, CAD, logistics, and other necessary and worthwhile team functions can get more of the attention they deserve. And we have made significant progress. However, our tendency to let build hog it all continues to challenge us. Like addicts, we have to be continually on guard against sliding back into old habits.

Without a Stop Build Day, we will tempted to keep building all the time and never stop.

However, one happy consequence of our efforts to kick the always-be-building habit has been developing a strong core of mentors and student leaders who are alert to the issue, and recognize its danger signs. I am cautiously optimistic.

6 Likes
#8

This year we spent too long on the build. While I think our robot turned out quite well, especially for a rookie team, (a lot of people were complementing us on how our robot looked) it caused our programming, driving, etc suffered as a result.

My point is, for the top teams, this isn’t gonna change much in terms of scheduling. The best teams will be the ones who can keep and stick to a schedule. And that schedule will probably still aim to finish the robot before the first six weeks of the season. We tried to do that but (perhaps due to our inexperience) we didn’t finish the build until March (we had a second robot). Yikes. Scheduling is definitely a huge priority for us going forward.

1 Like
#9

My name is Gus and I am a build-a-holic.

3 Likes
#10

We’re planning on keeping the same schedule. Finish two robots within 6 weeks, and use the practice bot to find breaking points. Fix issues on the competition bot accordingly. No bag will just allow us to install iterations to the competition bot before competition, saving a lot of time when we get there to allow for more practice on Thursdays.

#11

Footnote: this only works in about 7 states.

#12

There’s more out there than TBA would lead you to believe. For example, Iowa has a good Week 0 run by 525 that isn’t posted to TBA.

#13

Same, our team (7476) decided to build a rocket bot. It took so long that we basically had drive code, a camara, and a very bad PID. I think what is important is to not ALWAYS try to build. Sometimes priorities need to be made.

2 Likes
#14

FTFY

#15

I have two mantras that I express early and often:

  1. KISS - look for simplicity in everything you do. Simplest design, simplest manufacturing, lowest degrees of freedom. Do not get seduced by the dark side of being “clever” or different just to be clever or different. Reusing proven designs is usually better that starting from scratch. Steal from the best. Look for proven solutions to derisk your design and fab processes. Its engineering not art; there is no shame in reusing a proven design or process. Decide upfront what level of design complexity you can deliver and then stick to that budget. If you add complexity in one part of the design, complexity has to come out somewhere else. There is a reason that robots that can do it all are called “godbots”, and are only truly delivered by a very small number of elite teams. Allow time for software integration and drive practice in your schedule.

  2. Less is more - It is far far better to do less at 100% success rate, that attempt to do more and end up only doing it at 80%. Go with an iterative design process and add additional complexity once you have the basics nailed. You don’t have time for that many cycles during a match, you really need to them all to be successful. When in doubt, do the math!

#16

Ah, EOM Lions! I will say, you guys did incredibly well for your rookie year and I’m excited to see how you do next year now that you’ve had some experience under your belt.

#17

I suspect there will be a lot of “me-too” robots. FTC has been running with this model for several years now. By the time you get to Champs, almost all of the top tier robots are the same basic design concept. It will take a very creative GDC to make sure that one type of robot does not dominate the entire game.

#18

Do you think for 2017,2018, or 2019 there would have been a specific “me-too” robot? What do you think it would look like? Do you think teams would copy specific mechanisms or whole robots like we see in FTC?

1 Like
#19

I think you will see more copying of mechanisms. FRC teams tend to be very attached to their particular chassis configuration. That will provide some limiting factor for which mechanism make sense in combination with other parts of the robot such as drive train.

#20

Yeah, I don’t see copying of a mechanisms as a bad thing. Take a look at 254’s cube collector last year, a lot of teams had a version of it by world champs. The same could be said this year with the compliant wheel hatch mech. Imo we will see more successful versions of others teams mechanisms, but this will now be possible for teams without the resources to have a practice bot. If you had a practice bot this year, making a wheel hatch mech, installing it, and practicing with it on your practice bot was a huge advantage, that sadly teams without the resources to build two robots can’t obtain.

1 Like