Did you make a second robot? [POLL]

[DonRotolo]

tl;dr

The key to winning a competition is a capable robot and capable drivers.

Robots are capable when the design and fabrication processes used are good (arguably).

Drivers are capable when they have enough experience driving (altough some are naturally better than others). If your drivers show up at competition with less than about 20 hours of driving experience, you are at a disadvantage. If they have 100 hours, they can make up for a lot of 'bad' in a robot.

So, between bag day and competiton, what's your drve team doing? Sitting on their hands, or driving your practice robot?

I would be interested in knowing if any team that did not build a second robot has ever made it to Einstein. I'd be surprised if the number was much greater than zero.

[Michael Corsetto]

Originally Posted by DonRotolo

tl;dr

The key to winning a competition is a capable robot and capable drivers.

Robots are capable when the design and fabrication processes used are good (arguably).

Drivers are capable when they have enough experience driving (altough some are naturally better than others). If your drivers show up at competition with less than about 20 hours of driving experience, you are at a disadvantage. If they have 100 hours, they can make up for a lot of 'bad' in a robot.

So, between bag day and competiton, what's your drve team doing? Sitting on their hands, or driving your practice robot?

I would be interested in knowing if any team that did not build a second robot has ever made it to Einstein. I'd be surprised if the number was much greater than zero.

I think 610 is known for not building a practice robot? Was this the case when they won in 2013?

Given that 32 teams make Einstein now, I'd guess there are at least a few, if not more, that have made it to the big show with no practice robot.

-Mike

[BrennanB]

Originally Posted by Michael Corsetto

I think 610 is known for not building a practice robot? Was this the case when they won in 2013?

That is correct. They still do not build one.

[pilleya]

In 2015 we built two robots in sequence, but they were completed within a close timeframe of each other. We did not really decide which robot we would bag until the last week or two of build season.

This season we decided to not build a complete practise bot but pretty much possess an entire second robot+more in spares/parts. We spent a lot of time fine tuning our shooting mechanism during the build-season which used up a lot of CNC time, we built 3 different shooter designs with around 5 shooters built in total We mounted our shooter on a test bed which was the same dimensions as our chassis, to help our driver practise line the turret up for shots and also to help test and fine tune our vision tracking.

We are planning to build a practise robot next year.

[Thomas610]

Originally Posted by Michael Corsetto

I think 610 is known for not building a practice robot? Was this the case when they won in 2013?

Given that 32 teams make Einstein now, I'd guess there are at least a few, if not more, that have made it to the big show with no practice robot.

-Mike

Ya, we generally don't build a practice bot (as was the case in 2013), mainly due to the effort in manually machining two full robots. Needless to say, not having a practice bot has been annoying at times, but it certainly doesn't stop us from iterating and improving.

I will say that having a practice bot is really nice for testing and flushing out designs to see the weaknesses. We had to rush finishing our bot in 4 weeks so we could had the remaining 2 as practice/time to break the bot and improve designs.

We did make 2 bots in 2010 and just ran into a bunch of issues. Whenever we were making new autos, we found that whatever we did on the practice bot always required a lot of work to get working on the real bot (probably due our feedback integration). Additionally, we make lots of changes here and there on our real bot and it's really annoying to do that twice to truly have an identical replica. We still found the biggest issue was actually making two bots. It took us a week and a half to fully machine out bot this year, it's just unreasonable for us to spend a solid 3 or so weeks machining parts...

This year, we designed the backbone (the superstructure of the robot and the claw) of our robot to be easily removed (and lucky it was under 30lbs ). This lets us do maintenance and upgrades of our bot without having to make a practice bot. If you look at our robot between GTRC and ONWA, we managed to get our high goal shooting nailed down (pretty huge change in performance) by working on the real claw for our bot. Kind of a nice balance between making a practice bot and not.

Tldr: We don't build a practice bot because it takes too long to machine for a replica that never is truly accurate to the real thing. Instead, we built our bot to be modular so we can iterate on certain parts within the withholding allowance.

[engunneer]

Count us among teams that build two non-identical robots. the first one had a few dimension and construction issues, and we've built 3-4 complete shooter assemblies by now between pre-bag and various withholding allowances.

The second bot was made better, and used what we learned on the first, but they are similar enough that the code transfers with few modifications between them. Electronics mounting is another story, and the practice robot only has a front bumper because our intake requires one.

[hardcopi]

We have 2 bots that are VERY close. The competition bot though is so much better quality. The practice bot has never met Mr. Loctite and is always at the edge of flying apart. Has a ton of extra holes where we tried something, has 2nd hand motors, etc. It also has dents where we determined at what speed hitting the rock wall would actually launch us completely in the air landing with a scary sounding thud.

Poor thing is the abused older brother. The younger competition bot has new motors (just got all new drive motors), everything is loc-tited and only has the needed holes.

[hyperhawk]

Our team tried to make a practice robot this year, but the process for us was very much so disconnected from the building of the actual robot. We intended to build the chassis to practice driving and also to get a head start on the coding. For us, quite possibly since it was the first time with a practice robot, it didn't exactly seem to have a significant impact on the quality of our coding advances or drive training. We seemed to fall behind on the main robot and ended up not adding any of the additional components to the practice robot for testing or any of those things. It may have been a personal thing for our team, and I can definitely see how having the second robot could be better in years to come as we get more experience with it. Some ways that I can see that it could have significantly helped us would be that we could have worked out our climbing mechanism (which was not used until our second competition), while working on other parts of our competition bot. It almost seems as though we forgot that we had another robot. It seems that the most difficult thing about having the second robot would be managing the time around it.

tl;dr Practice bots are a great idea in theory as long as efficient time management is present

[nuclearnerd]

Last year we built a second chassis from an old c-base kit and withheld our mechanism to get a practice robot on the cheap. It worked out really well, but we spent a lot of time during the first day of regionals putting our robot back together.

This year we built two identical robots, and it helped. But when I look at our bills, and I read through this thread, all I can think about is how *incredibly wasteful* all these duplicated robots are! I know the bag rules have been debated to death, but I hope something changes in future to let teams be competitive without having to double their budget.

[TheMilkman01]

We called our practice robot "Bubbles." How that goes with our official robot name "Termite" I do not know.

[Anupam Goli]

We've built practice robots every year since 2013, and found that it has helped immensely in doing design iterations and driver practice.

This year, we built two identical chassis, frames, and intake mechanisms. However, we built only 1 shooter module that we would use our withholding allownace on. We did this so we would only have to tune one shooting mechanism, and we wouldn't have to waste time retuning the competition version when we unbagged the robot.

[philso]

Originally Posted by Thomas610

This year, we designed the backbone (the superstructure of the robot and the claw) of our robot to be easily removed (and lucky it was under 30lbs ). This lets us do maintenance and upgrades of our bot without having to make a practice bot. If you look at our robot between GTRC and ONWA, we managed to get our high goal shooting nailed down (pretty huge change in performance) by working on the real claw for our bot. Kind of a nice balance between making a practice bot and not.

Did you mount the backbone/superstructure on something like a chassis from a previous year so you could drive it around and simulate it being on your "real" chassis?

[philso]

Originally Posted by Type

We built a practice robot that uses similar controls to our robot and similar function but looks entirely different and drives different. It's essentially just to test code, like we used it to figure out how to use two autos, then we also used it to work on vision tracking. It's not as good as our actual robot, but it's better than nothing.

Originally Posted by engunneer

Count us among teams that build two non-identical robots. the first one had a few dimension and construction issues, and we've built 3-4 complete shooter assemblies by now between pre-bag and various withholding allowances.

The second bot was made better, and used what we learned on the first, but they are similar enough that the code transfers with few modifications between them. Electronics mounting is another story, and the practice robot only has a front bumper because our intake requires one.

Originally Posted by nuclearnerd

Last year we built a second chassis from an old c-base kit and withheld our mechanism to get a practice robot on the cheap. It worked out really well, but we spent a lot of time during the first day of regionals putting our robot back together.

It seems that most teams that do build practice robots make them fairly close to the competition one.

For those who build practice robots that are not close to identical to the competition one, what are the pros and cons for the practice robot being different? What are the major differences (other than cosmetics)? Would you continue to build practice robots that are different from the competition robot?

[Type]

Originally Posted by philso

It seems that most teams that do build practice robots make them fairly close to the competition one.

For those who build practice robots that are not close to identical to the competition one, what are the pros and cons for the practice robot being different? What are the major differences (other than cosmetics)? Would you continue to build practice robots that are different from the competition robot?

So we have a fairly rookie team this year. There are only 2 students returning so we tried to make a T shirt robot to learn more about building a robot before season. We used last year's KoP to do it. We never finished it so we decided to use it as a practice robot. Currently, we are using rhino tracks for our competition robot. To move our shooter on our arm up and down, we have 2 dart linear actuators ran with cims, our practice robot has an entirely different type of shooter setup. Our little shooting basket that would be impossible to load during a match is ran by a window motor and a threaded rod basically. Our practice robot can't do defenses, can't shoot as accurate, and doesn't even look close to any robots you will see this year but we use it to test code . We have rookie programmers so we use it to figure out how to do things. Next year I'm hoping we can make an identical replica so we can practice driving and code. Most of our improvements we design in CAD so we don't really need a practice robot in that aspect.

[Thomas610]

Originally Posted by philso

Did you mount the backbone/superstructure on something like a chassis from a previous year so you could drive it around and simulate it being on your "real" chassis?

Ya, that's exactly what we do. We've just made an adapter that holds it in the same rough position as on our real bot.