How do you make your second robot?

[rvgrossman]

For teams out there that have an identical second, or even third, robot. How do you build your second bot? Do you build both at the same time or do you do one and then the other? On top of that, any tips for a newer team trying to do it next season?

[Richard Wallace]

Our team built two (nearly) identical robots for the first time this year.

I think there were three key factors that made it possible:

1) we had sufficient resources (people, skills, space, & funding)

2) our design was guided by strategic priorities*, and

3) we built the two robots simultaneously, and did not decide which one was which (competition 'bot or practice 'bot) until bag day. Strategic priorities drove that decision.

----------
*Our process for developing strategic priorities would require a longer discussion. We borrowed heavily from approaches published by successful teams to find our way.

[troy_dietz]

Originally Posted by rvgrossman

For teams out there that have an identical second, or even third, robot. How do you build your second bot? Do you build both at the same time or do you do one and then the other? On top of that, any tips for a newer team trying to do it next season?

From what people have explained to me and from personal experience, we (330) build ours more or less at the same time, but during the last few days of build season all efforts are focused on the competition robot. (If it has not yet been completed) We usually spend several weeks designing and are left with little time to build. (in comparison to other teams)
The 2nd robot is usually built using parts from previous years robots. (practice AND competition if necessary)
This year, our practice field was built throughout the season and was finished in the last few weeks. (The elements weren't fully arranged and taped out until the first weeks of practice after B&T)
If you don't build a practice field already and plan to build one for a practice robot, I HIGHLY recommend constructing it as soon as possible (within the first week of the season). If you can spare non essential personnel from the design and planning team, have them start on it.

One more thing: If you paint/anodize/powdercoat your robot, do that to any removable mechanisms on your practice robot as well. (object manipulator, endgame mechanism, etc)

[EricH]

We built our practice robot first. That showed what didn't work properly.


Then we made upgrades as we built the competition robot.

[Clayton Summerall]

This year we decided to make a 090 bent sheet metal robot. This being said it if we could figure out how to do this for the first time, it would be easy to have a second set made. So in week 5 our sheet sponsor came through with 1 1/2 robots worth of our parts, this lead to us to painting the bot in the back yard, then the next day assembling. We decided to make the first one the flight as we were unsure if they were going to get the second one done in time. So after the build season was completed we put the second bot together, something like week one regionals, we had the second bot finished. Although this process was stressful, it was nice having 2 robots, as the practice was crucial.

[evoluti1]

Originally Posted by EricH

We built our practice robot first. That showed what didn't work properly.


Then we made upgrades as we built the competition robot.

I'll elaborate: prior to this year, we never built our practice bot with the explicit intention of making it identical to the competition robot. That is, the two machines would have similar mechanisms, but the practice bot might have components made from wood instead of aluminum, it might use illegal motors, pneumatic cylinder geometry might be different, etc.

The problem with that approach is that you actually have to design two robots. If you don't make a concerted effort to keep the two machines as similar as possible, you will have to solve an entirely different set of problems for each robot to get it to work. And any amount of time you spend solving problems on your practice robot that don't apply to your competition robot is time you're not using to make your competition robot better. This defeats one of the main benefits of building two robots. It might be counterintuitive, but you build a better robot and save time if you go straight to designing your competition robot, rather than spending too much time on a "prototype" version first.

Having learned this lesson the hard way, we went into build season this year with the intention of building two robots to be basically identical. One of the main problems our team struggles with is consistent craftsmanship, and a key way we mitigated that this season was by making corresponding parts for each robot at roughly the same time. For the most part, we would build a part or mechanism for the practice robot, and after verifying that our methods were sound, we would immediately build a second copy for the competition robot. This helped ensure the same sets of hands were responsible for both copies of a part, which in turn helped ensure uniformity between the two machines.

In reality it didn't always go as smoothly as that, but those are the basic details. I think I can safely say having a nearly-identical practice robot was a big factor in making this our most competitive season so far. If you have any other questions I'd be happy to tell you more about our process.

[Pretzel]

Originally Posted by rvgrossman

For teams out there that have an identical second, or even third, robot. How do you build your second bot? Do you build both at the same time or do you do one and then the other? On top of that, any tips for a newer team trying to do it next season?

Team 1619 has come up with a way of building two robots that is slightly less resource/time intensive as building two identical robots. We build two identical drivetrains, and then only one of our primary manipulator which we then use as our witholding allowance.

Since our changes between events have never involved changes to the drivetrain, we are able to ensure that the manipulator that worked in practice is the same one we use on the field.

In 2014 our catapult and collector were mounted together onto two plates that attached to the robot using 8 bolts. To remove the whole assembly we would unplug two pneumatic tubes and two electrical connections (for a laser sensor and a motor) before just unbolting the 8 bolts and lifting it off to take with us to competition. Re-installation was similarly a breeze.

In 2015 it was a bit more complicated, simply due to the mechanisms and how we wired it to best conceal the cables. We had a giant claw, used to pick up containers, that we would remove to swap between the practice chassis and the competition robot. We also had a conveyor belt that sat in the middle of the robot that we had to replace at our first competition for weight savings. The biggest hassle there was the electrical connections. We could take the entire "power tower" (the claw on it's long rail that allowed it to travel up and down) off the robot with 3 bolts, but wiring/unwiring the 2-3 motors (depending on which regional it was for), the encoder, and the limit switches took a while since the whole bundle was neatly enclosed in snakeskin and ziptied down. It was still a much better solution than if we had created two full robots however.

If you create two of your drivetrains and your manipulator design is such that it weighs less than the witholding allowance, it cuts down on the majority of the different revisions of machining you would need to do if you had two identical robots the whole season. You only need to keep making stuff to account for the changing design of one manipulator, not two, and the drivetrain usually doesn't change much from competition to competition.

[hardcopi]

We built the practice bot first. As we did though we were building the parts for the second. Then day or two before bag day it was jigsaw puzzle time. Put it all together.

[GeeTwo]

We have a subteam working each subsystem. This usually means that we build one instance, and then the other. We don't decide until within a few days of bag which one will be the competition robot and which will be the practice robot.

Our drive chassis the past two years have been mods to the AM14U2, which makes it easier to keep them identical, and provides lots of standard mount points for the manipulators.

For Recycle Rush, we changed our mind as to which robot would be bagged within 24 hours of bag as the second robot finally came together.

For Stronghold, we had two identical drive chassis (including controls and pneumatic infrastructure), two identical launchers (although one of them has not been final-assembled as our pneumatic cylinders did not come in until after bagging), and several pickups (five) that were each improvements on the preceding design. The first four iterations became better at pickup and delivery to the launcher, the final is not quite as effective at pickup as the preceding model, but is significantly more robust.

[MrForbes]

we did our first practice robot this year....it was relatively easy and inexpensive because of how we built our competition robot, we were able to re use last year's chassis for the practice bot.

Generally, it's nice to get the practice robot done first, then add the 5th and 6th week improvements to the competition robot before you bag it. actually doing it this way can be challenging, since the improvements never stop, so you have to manage the project carefully and set a cutoff time.

right? I'm just going from what I've seen other teams doing...

[jnicho15]

548 makes a practice bot in the first 3 or so weeks to try out design ideas. Then, in the last 3 weeks, we refine the design and build the comp robot. After bag day, we make practice identical to comp.

[mrnoble]

This year was tough for us. 1339 designed and built a practice bot and competition frame (ready to powder coat), but then after testing decided that the drive train wasn't adequate (it was a cantilevered brecoflex tank). We went through a complete redesign of the drivetrain in week four, so we ended up building a new competition not, then retroactively building a new practice bot as we approached bag and tag. We won't make that mistake again, I hope.

[seg9585]

Team 4276 built 2 identical chassis and test-fit components on the practice chassis while the other was getting powder coated. Once we were happy with the component positions on the practice chassis, we drilled holes and mounted to the final robot.
Ultimately, the practice robot ended up with significantly more holes in it (as we changed positions of parts). We completed the competition robot in time for Week 0 scrimmages, then bagged it up and spent an additional week completing the practice robot so the drive team could practice with it.

At this point, our practice bot has been partially cannibalized for replacement parts on the competition bot and to test upgrades. We intend to fix it back up so we can run both at demos after CMP.

[BumblingBuilder]

For the most part, we build the two bots simultaneously. We fabricate the parts, and the less desirable parts go on the second bot and the better parts go to the competition robot. In assembly, unless plans change (usually time constraints caused by a variety of potential problems), the practice bot is always a little ahead because it is on the practice bot where we address potential problems and fix them so we don't have to face them on the competition bot. It becomes both a practice bot and a design debug robot.

However, this year, we made many flaws in design (not too mention design did take a while). Enough so that, for a time, we ditched the practice bot and worked solely on the competition bot because we worried about how much we could actually get done before bag. This scramble to finish and address problems has become a great learning experience for our team and we will make sure we try not to repeat them (although I could argue many of the flaws we made this year resemble the same ones we made in 2015, so we hope not to repeat them again. Trust me we are not an example of Einstein's (?) definition of insanity).

[Eric Scheuing]

We usually build our "second" robot first. That allows us to troubleshoot as we build, and come up with some more elegant and robust solutions to any issues we come across for the main bot. Usually we'll fabricate the parts for the main bot as soon as we have a workable solution found for whatever issues we come across, and assemble as needed.

Of course, if we're having any parts made out-of-house, we fab them all at the same time. It would be a huge waste of tooling time to have someone make the parts for us only to have to make them again a couple weeks later.