Looking Back: 3 Day Robots

[aldaeron]

At the beginning of the season when multiple "3 day robots" (Ri3D, BuildBlitz) were announced, there was a lot of discussion on whether or not it was a good thing or not to have these televised, "professionally built" FRC robots.

Here is one of the many threads: http://www.chiefdelphi.com/forums/sh...d.php?t=123152

I setup a reminder to follow up and now I want to see what you all think. I am not sure I have finished making up my mind, but here are my initial thoughts.

The biggest thing I wonder is what designs we would have seen without the 3 day robots? Would teams have come to similar conclusions in 6 weeks that the 3 day robot teams did in 72 hours? Would we have seen different ideas at the competitions if there were no 3 day robots?

I admit it was a little disappointing to see some bots that looked identical to what I saw on day 3 of the build season on YouTube at the competitions. Listening to my scouting team describe teams was often similarly disappointing. Some things I heard were: "it's a SimBot with a JVN intake" and "it's literally a Boom Done clone".

I never was able to investigate these teams in more detail in the pits. Maybe they were a small team with limited financial resources and this was their first solid bot that could play the game well. Perhaps they were rookies with limited professional engineering mentors and would have struggled to do more than driving. Maybe this was the first year a team was able to have a robot done by week 4 and actually spent time testing or doing driver practice.

Many teams made clever combinations of the 3 day robot designs, combining their strengths and improving on their weaknesses.

Overall I thought the 3 day robots added to the season and would like to see them come back. I think that releasing all the code and CAD after the 72 hours was a good thing because it gave teams a working example to look at.

My main dislike was that the Boom Done team kept working on their bot and releasing updates during the season (I do not think any others kept working, but I could be wrong). While it was interesting to see their progress, I think that limiting the FRC professionals to a short amount of time helps showcase good concepts, while not figuring out all the fine details of a design for the students. The amount of work these FRC veterans can crank out in 72 hours is amazing and a huge help to many teams.

The opinions above are my own and are offered in the spirit of healthy reflection and debate. It is not my intent to diminish anyone's design or critique how their team operates.

-matto-

[Max Boord]

I am a strong supporter of the Ri3D/ BuildBlitz robots for many perspectives.

1. They are a good example of what a mid level robot must/ will be capable of at regionals.
2. They provide a multitude of solid, tested, and documented mechanisms available from the start. This helps out teams who do not have the time, resources, or wish to prototype every type of mechanism capable of completing a task.
3. It allows teams strategy department to get working immediately knowing how the game will be played and what basic principals will or wont work.
4. For rookie teams, what better way to get an introduction to FIRST that seeing robots preforming this years game in a matter of hours.
5. The Ri3D builds can be an emergency backup in the event a team finds out late in the season there mechanism does not work.
6. By having 5 builds the chances of your alliance having a toaster/ printer/ box on wheels robot goes down significantly.
7. By showing off low cost, easy access machinable parts (versa system) less machining capable teams will have a solid place to start on construction materials and techniques that actually apply to them.

[dudefise]

I think it's a good thing, as others have said, it provides options for teams unable to design their own functioning mechanisms due to delays, and helps those who would otherwise be competing with box-bots. This increases the competitiveness of the competition, which helps everyone.

The downside is that it decreases creativity, with many designs from teams capable of designing and constructing unique components simply being copies of the Ri3d stuff.

Our team was able to counter this effect by asking the students to avoid watching Ri3d until initial brainstorming, strategizing and conceptual design had been completed.

[BBray_T1296]

I liked the Choo-choo mechanism. Unless teams like us were looking at (EDIT: I knew better) Winnovation 2008 robot (or people who worked on it ahem Aren), I doubt that many teams would have stumbled across this great mechanism.

[nuclearnerd]

Originally Posted by BBray_T1296

I liked the Choo-choo mechanism. Unless teams like us were looking at Simbotics 2008 robot (or people who worked on it ahem Aren), I doubt that many teams would have stumbled across this great mechanism.

It's very cool, no doubt, but if you have access to the other side of the pivot, I think a snail cam is a better equivalent. Much easier to build strong without so many cantilevered shafts.

Back on topic - I'm a supporter of RI3D and Build Blitz. Sure we've seen a few hundred JVN low pivot winched catapults, O-Ryon claws with kickers, and BoomDone/Copioli motorized high pivot catapults this year. But I'd rather see what teams can do with that inspiration than a bunch of schools flounder looking for an idea. (Much of the similarity is probably due to design convergence, rather than imitation anyway)

[Chief Hedgehog]

The jury is still out for me.

After attending a week zero event and then having watched countless hours of video via TBA, followed by our lone regional at North Star I keep seeing robots that were near mimics of the Ri3D Bots. After all of this, I am left wondering if we would have seen more differentiation of designs in the robots themselves without Ri3D.

I give all credit to my design team that was lead by Ginger Power by creating a Robot that was very different in design and game play in reference to the Ri3D Robots. Our robot had a very unique design structure that included a ground floor pick-up via a fork design (which we did not utilize at North Star). Our throwing mechanism - a catapult with a very unique cam that was powered by a winch that my students built and stressed with speargun tubing http://www.magisto.com/album/video/f...8wNSIHDmEwCXl9

What you do not see in this video is how the lead screw has put tension on the tubing and then the student-built winch pulls it back into position. But if you watch our videos from North Star, you can see how effective it can be - and during this event we were only pulling back at around 40% - if we pushed it the full 100%, we would be throwing outside of the HP area with ease. We never pushed the limit as we saw no need to. Especially when we were drafted by the #1 alliance to keep doing what we were already capable of.

We struggled early in our Regional as we were not able to fit the Superstructure to the competition robot - but once we did, we were dominant.

I see the positives of Ri3D, but I also see how it can severely limit the ingenuity of students - especially if the teams are heavily influenced by the Ri3D designs, or by the mentors that are swayed by Ri3D.

I know that I may be a little out on the fringes as I allow my students to make the decisions of the final design (but we have been fairly sucessful the last two seasons) - but as a coach and a teacher, I realize the valuation in failure. And in the last two seasons, 4607 has failed a lot in the initial design phases of our robot (any inspector at North Star in 2013 can attest to this as we were the last robot to pass inspections - where we had to rebuild our robot on the spot).

But I believe in my students, I encourage them to fail in their endeavors to find the best formula. There is valuation in failure. That is why you will never see me with the drive team. I feel that my coaching has already been done before the regional - I have no need to be on the drive team - just like I have no need to be on the field when my soccer players are playing the game.

Aside from all of that hyperbole, I do see value in modeling. If a team can gain insight from Ri3D and then improve, more power to them. The competition of ideas and ideals is why FRC has quickly become my favorite of all sports - no matter of the controversies.

[MrForbes]

Originally Posted by aldaeron

I am not sure I have finished making up my mind

Originally Posted by Chief Hedgehog

The jury is still out for me.

Me too. I can't vote on a poll that appears to be a false dichotomy. How about you add a selection for me?

[Anupam Goli]

I suppose my opinion has been fully formed now that my team's season's over. I've been to two competitions, watched many others, and must have seen hundreds of robots by now. A lot of robots look similar to each other, and a lot of teams have "clones" of Ri3D and build blitz robots, but all of these teams went through a journey to build their robots, and I'm sure that along the way these teams had to iterate their designs. I don't think it's possible to put together a Ri3D/BB "clone" without learning something or having to iterate a part of the design to match your resources.

Not all robots that look like the 3 day projects copied them. Plenty of high performing teams look similar to the 3 day projects, but most likely they prototyped many different mechanisms and through iteration developed their current robot. This may have been the most successful year my team has had, and also one of the most complicated robots we've ever built, but if you strip away the catcher walls, it looks like an 8020 and versaframe version of a team JVN build. (For Reference). I met with an old mentor of mine, and I showed him our robot. His first response was "You guys copied the robot in 3 days, didn't you?". I explained that while our intake and catapult look very similar, the amount of prototyping, designing, and engineering that went into those systems took more than 3 days. I was actually inspired by 2012 over-the-bumper intakes when I first sketched up what would end up becoming the intake on our robot. We did use the "choo choo" mechanism that Aren Hill came up with, but we must have gone through at least 7 iterations of the darned thing before we finally developed one that worked with our system.

I think these 3 day builds have done a great job in helping teams get started and have certainly reduced the number of robots incapable of manipulating the game pieces. Not every robot can manipulate game pieces well, but that's solved by iterating and improving designs, which falls onto teams. It's not necessarily a bad thing that robots look alike. Sure, there's a creativity award, but I think every team wants to field a robot that works, and when there are 6 different concepts that are already presented, it's much easier to do that than starting from scratch.

[Dave Campbell]

Another point to keep in mind - teams do not have to look at the Ri3D videos and information. It can and should be a team decision to figure out what references and sources they use for ideation and robot design. Ri3D robots are just another source teams can use or ignore.

[cxcad]

Honestly, I think the 3i3d has little effect on the design on more established teams. There are only so many "obvious" designs out there; look at 2012, a lot of robots looked the same and so did in previous years. I think the only thing 3i3d did for the more established teams was verifying concepts

[Mastonevich]

I think overall they add value to the program.

While it may lessen the creativity for some, it seems to bring up the floor for a greater number. Even with the designs there were many "box on wheels". It is still not easy to recreate for many teams.

I believe more students are inspired by a better competition with the floor being raised, as opposed to the less students inspired with the lowered creativity.

It appears to me to be a net overall win for the program.

[KevinG]

I think our team took too much inspiration from Ri3D. Next year I think we're going to try and sit down and work out the decision matrix weightings before we look at what other teams are doing. There's nothing wrong with seeing how something is successfully implemented and getting inspiration from it, but there's something to be said for having the students design something on their own.

Ultimately it's another resource, no worse than CD or looking at youtube videos of previous years. There's only so many ways you can handle a giant inflatable ball, just like last year there were only so many ways you could throw a frisbee.

[MrBasse]

Originally Posted by Dave Campbell

Another point to keep in mind - teams do not have to look at the Ri3D videos and information. It can and should be a team decision to figure out what references and sources they use for ideation and robot design. Ri3D robots are just another source teams can use or ignore.

But they do, so many teams do, and we saw so many duplicates this year. I like the additions that Ri3d/Build Blitz can add, but why not do them in week 2 or 3? Still give plenty of time to borrow ideas, but also gives teams a chance to solve their own problem first. Lots of my students saw the solutions developed by the 3 day teams and those were the only possible solutions they saw all season. It effectively removed a few of our students from the unique and creative design of our robot. High School teenage tunnel vision is a powerful thing...

[Anupam Goli]

Originally Posted by MrBasse

But they do, so many teams do, and we saw so many duplicates this year. I like the additions that Ri3d/Build Blitz can add, but why not do them in week 2 or 3? Still give plenty of time to borrow ideas, but also gives teams a chance to solve their own problem first. Lots of my students saw the solutions developed by the 3 day teams and those were the only possible solutions they saw all season. It effectively removed a few of our students from the unique and creative design of our robot. High School teenage tunnel vision is a powerful thing...

In this case, it might be the responsibility of us mentors to make sure our students aren't as affected by Ri3D. A lot of teams that copy Ri3D and build blitz can't get it working very well in the full 6 weeks. I'm not sure if having them release a good 2-3 weeks into the season will help at all. It will probably hurt most teams who will immediately switch design, wasting the time before Ri3D and Build Blitz reveal.

[Andrew Schreiber]

Originally Posted by BBray_T1296

I liked the Choo-choo mechanism. Unless teams like us were looking at Simbotics 2008 robot (or people who worked on it ahem Aren), I doubt that many teams would have stumbled across this great mechanism.

Here's a fun story, I've been around a while and I knew about the 1625 choo choo from when it was done in 2008. However, I had a student, who watched Team JVN, explain it to me ... this was the first time I understood how it worked. So, inspiration managed.


As an addendum - 1114 did not use that linkage in 2008. They used a dog released winch.