Creating the perfect robot

[MichaelBick]

Originally Posted by Thunder910

Ability to full-court shoot very accurately and limit the level of defense to be played on them- High release point, accuracy, fast lineup (either by code or a hard stop)

Honestly I don't see why a perfect robot needs to have an FCS. At the top levels of play the field is going to be deep enough to run a 3 cycler alliance(which as we well know, can pretty reliably clear out the feeder station). In a shallow field you can beat strong alliances regardless of FCS. If you still decide to run an FCS, there should be no reason why it has to be tall. If you are only FCSing you will promote defense. A much better style is a run and gun style similar to that of 469.

[Kevin Leonard]

Originally Posted by MICHAELABICK

Honestly I don't see why a perfect robot needs to have an FCS. At the top levels of play the field is going to be deep enough to run a 3 cycler alliance(which as we well know, can pretty reliably clear out the feeder station). In a shallow field you can beat strong alliances regardless of FCS. If you still decide to run an FCS, there should be no reason why it has to be tall. If you are only FCSing you will promote defense. A much better style is a run and gun style similar to that of 469.

What if you get picked by a good floor-pickup like 33 whose cycling isn't all that great?
What if you're completely outgunned in a cycling battle in a qualification match?
Full-court shooting is a game-changer. It forces your opponents to alter their game strategy to deal with your full-court shot, and if they can't block you, they likely lose, especially if you can then pick up your own misses and shoot those. Even 2-point full-court shooting is deadly provided those discs aren't going to be scored otherwise.

Ex1: If you DON'T block 67, you lose. They can score over 200 points on their own.

Ex2: 2169 in Galileo is arguably the best dedicated full-court shooter in the world. Their FCS game is so #solid that they were the only teleop threat (that was leveraged) on their divisional finalist alliance. Their FCS shot was also difficult to block because of it's high release point.

[evanperryg]

What if you get picked by a good floor-pickup like 33 whose cycling isn't all that great?

We are talking about an all-around, god-tier robot. That robot would be the one picking, if anything.

Anyway, I would like to inform all of you that I am now CADing the ultimate ultimate ascent robot. Might post pics if I can finish it.

[Kevin Leonard]

Originally Posted by evanperryg

We are talking about an all-around, god-tier robot. That robot would be the one picking, if anything.

Anyway, I would like to inform all of you that I am now CADing the ultimate ultimate ascent robot. Might post pics if I can finish it.

I'd like to point out that the three best robots in Archimedes got picked, they didn't captain their own alliances (i.e. 254, 469, 987).

If this robot couldn't stand up to a match against three good cyclers with limited to zero help (i.e. being able to FCS, as that's they only way they would stand a chance), then they wouldn't sit at the top of the rankings, as they would lose that match.

And I totally want to see this CAD. It'd be epic.

[evanperryg]

Originally Posted by Thunder910

I'd like to point out that the three best robots in Archimedes got picked, they didn't captain their own alliances (i.e. 254, 469, 987).

If this robot couldn't stand up to a match against three good cyclers with limited to zero help (i.e. being able to FCS, as that's they only way they would stand a chance), then they wouldn't sit at the top of the rankings, as they would lose that match.

And I totally want to see this CAD. It'd be epic.

A fcs with no help vs 3 cyclers would get pounded, too. Also, the base is almost done. Making a request for 67, 987, 254, 1114, 2056, 610, 1310, and 3476 to post their CAD.

[Kevin Leonard]

Originally Posted by evanperryg

A fcs with no help vs 3 cyclers would get pounded, too. Also, the base is almost done. Making a request for 67, 987, 254, 1114, 2056, 610, 1310, and 3476 to post their CAD.

I think an undefended 67 would crush most 3-cycler alliances.

[BobbyVanNess]

I've done a lot of thinking about this on my own, especially after IRI. I define the "Perfect Robot" as the bot that would most likely go undefeated through official match play, all the way to Einstein. In my mind, this would require the ability to reliably come somewhat close to producing the absolute maximum theoretical score for one robot. There will always be matches where the odds are against you, which means you'd need to have the greatest odds of winning any match, requiring a lot of versatility.

So let's start with the autonomous. The most we've seen from one bot is 7 discs (8 from 254 picking up misses), but I believe it might be faintly possible for a team with a good enough floor collector and strong programming to score 8-9 discs in auto.

Then there's Tele-Op, where I believe it should have the capability to full court shoot when undefended, and empty the alliance station. It would then go across the field and clean up any of its own misses. Essentially, it would have the capability of scoring every one of the alliance's discs, and possibly then some off of the ground.

Finally, it would load up the colored discs (either from the alliance station or the ground) and climb to 30, scoring the discs in the pyramid goal. Depending on the speed of the full court shooter, and the ground load time of the bot, the time it takes to climb may not be a crucial detail.

Now the fun part: What IS this perfect bot? Maybe a climb like 25's or 842's could be implemented on a 469 type bot. From the way I've seen their collector work, it looks like the >7 disc auto could maybe be possible. However, the bot that I believe is closest to this ideal "Perfect Bot" is 67. Now if a bot like that could implement a ground collector, possibly even for autonomous only, their max score would be 7 discs in auto, 47 in the high goal during tele (45+2 alliance colored discs), 4 in the pyramid, and the 30 point climb. That adds up to 233 points. If a 9 disc auto were possible, then that figure would increase to 245.

Of course, is any of this useful as hindsight? Will it help us in the future? Absolutely not! But it certainly helps hold us over until next build season

[Kevin Leonard]

Originally Posted by bobby5150

I've done a lot of thinking about this on my own, especially after IRI. I define the "Perfect Robot" as the bot that would most likely go undefeated through official match play, all the way to Einstein. In my mind, this would require the ability to reliably come somewhat close to producing the absolute maximum theoretical score for one robot. There will always be matches where the odds are against you, which means you'd need to have the greatest odds of winning any match, requiring a lot of versatility.

So let's start with the autonomous. The most we've seen from one bot is 7 discs (8 from 254 picking up misses), but I believe it might be faintly possible for a team with a good enough floor collector and strong programming to score 8-9 discs in auto.

Then there's Tele-Op, where I believe it should have the capability to full court shoot when undefended, and empty the alliance station. It would then go across the field and clean up any of its own misses. Essentially, it would have the capability of scoring every one of the alliance's discs, and possibly then some off of the ground.

Finally, it would load up the colored discs (either from the alliance station or the ground) and climb to 30, scoring the discs in the pyramid goal. Depending on the speed of the full court shooter, and the ground load time of the bot, the time it takes to climb may not be a crucial detail.

Now the fun part: What IS this perfect bot? Maybe a climb like 25's or 842's could be implemented on a 469 type bot. From the way I've seen their collector work, it looks like the >7 disc auto could maybe be possible. However, the bot that I believe is closest to this ideal "Perfect Bot" is 67. Now if a bot like that could implement a ground collector, possibly even for autonomous only, their max score would be 7 discs in auto, 47 in the high goal during tele (45+2 alliance colored discs), 4 in the pyramid, and the 30 point climb. That adds up to 233 points. If a 9 disc auto were possible, then that figure would increase to 245.

Of course, is any of this useful as hindsight? Will it help us in the future? Absolutely not! But it certainly helps hold us over until next build season

Possibly with a 2474 style collector. It collects both forward and backward.
Maybe the shooter can be elevated to 60" to shoot full court, and lowered below 30" to shoot auto. I kinda like that idea. A shorter 67 with a 2474 collector. And MAYBE the ability to change their shooter height.

[chmorroni]

Just a thought, but I know our robot (a more-or-less FCS) and a few others that I saw had to let their shooter wheels speed back up after each shot. What if it had a system of 2 shooters, so it could go through all the alliance's white Frisbees much more quickly? Then you have more time to do each of the other things on the list.

[GarrettF2395]

Originally Posted by chmorroni

Just a thought, but I know our robot (a more-or-less FCS) and a few others that I saw had to let their shooter wheels speed back up after each shot. What if it had a system of 2 shooters, so it could go through all the alliance's white Frisbees much more quickly? Then you have more time to do each of the other things on the list.

This is exactly what I was thinking!

It sounds crazy but what if you had an elevating top shooter similar to 1477, and a semi-stationary bottom shooter (it would only adjust its shooting angle)?
That way's it would be possible to fire all of your alliance discs very quickly, then if you had a 30 point climb or a floor intake, you would have enough time to make good use of those design aspects.

You could elevate your top shooter to score the alliance discs into the pyramid goal for the full 30 points if you had enough time for 2 trips.

Now, lets say that you're a wizard of compacting your designs.
When your top shooter is in its stored position, you can fit under the pyramid, and your bottom shooter can receive discs from your intake system (again similar to 1477).

So now all we are missing is the 30 point climb.
Suppose you made the elevator system strong enough and tall enough to act as a lifting arm for a 30 point climber similar to the 1986 or 254 style of climbing.
You now have every aspect of of the game covered with one robot AND enough time created by the double shooter to do those aspects.

Obviously this is a ridiculous idea, but it is certainly fun to think about!

[Rynocorn]

Just stick the top half of 148 or 1806 onto the bottom half of 1310!!
That should work great right?

[Kevin Leonard]

Originally Posted by GarrettF2395

This is exactly what I was thinking!

It sounds crazy but what if you had an elevating top shooter similar to 1477, and a semi-stationary bottom shooter (it would only adjust its shooting angle)?
That way's it would be possible to fire all of your alliance discs very quickly, then if you had a 30 point climb or a floor intake, you would have enough time to make good use of those design aspects.

You could elevate your top shooter to score the alliance discs into the pyramid goal for the full 30 points if you had enough time for 2 trips.

Now, lets say that you're a wizard of compacting your designs.
When your top shooter is in its stored position, you can fit under the pyramid, and your bottom shooter can receive discs from your intake system (again similar to 1477).

So now all we are missing is the 30 point climb.
Suppose you made the elevator system strong enough and tall enough to act as a lifting arm for a 30 point climber similar to the 1986 or 254 style of climbing.
You now have every aspect of of the game covered with one robot AND enough time created by the double shooter to do those aspects.

Obviously this is a ridiculous idea, but it is certainly fun to think about!

But if you have 67's shooter, 195's shooter, or 2169's shooter, you have awesome accuracy AND zero-spin-up time.
I don't know if 67 and 2169 used code to keep the speed of their shooter or some other method. I DO know that 195 used an exorbitantly large wheel so that after every shot, the weight of the wheel would keep it spinning quickly.

[chmorroni]

Originally Posted by Thunder910

But if you have 67's shooter, 195's shooter, or 2169's shooter, you have awesome accuracy AND zero-spin-up time.
I don't know if 67 and 2169 used code to keep the speed of their shooter or some other method. I DO know that 195 used an exorbitantly large wheel so that after every shot, the weight of the wheel would keep it spinning quickly.

Our shooter was a 180 degree circular shooter with an 8" pneumatic wheel direct mounted to a CIM. It takes at least .5-1 seconds to speed back up after pushing a Frisbee through it. I think one way to avoid that might be to decrease the contact time of the Frisbee with the wheel.

[SoftwareBug2.0]

Originally Posted by chmorroni

Our shooter was a 180 degree circular shooter with an 8" pneumatic wheel direct mounted to a CIM. It takes at least .5-1 seconds to speed back up after pushing a Frisbee through it. I think one way to avoid that might be to decrease the contact time of the Frisbee with the wheel.

I don't think that how much energy you need to push the frisbee through really matters. You can almost always just add more power to get a faster recovery time. We had a similar setup to yours: an 8" Colson on a ~120 degree path. We started out with a CIM but then went to dual fisher prices to get it back up to speed faster.

I think it would be possible to get the throughput of this sort of shooter up above what a single human player can load without too much extra trouble. I think the ultimate FCS would need to take disks from multiple human-loading slots more than it would need a second shooter.

[cadandcookies]

Originally Posted by Thunder910

I don't know if 67 and 2169 used code to keep the speed of their shooter or some other method.

2169 had a hall effect sensor to keep track of their RPM and how quickly they could feed, in addition to having plenty of power and I believe a heavy wheel (the only thing that is definitely correct about that is the hall effect sensor).

The other thing about 2169's shooter is that it's angled so they could sit directly against the feeder station and still shoot "straight." If I remember correctly this affected where they could sit in autonomous and they had to be angled (which may or may not be an issue if you're repurposing it for pickup).