[FTC]: Robot Capabilities

[jamie_1930]

We're starting to near the last section of build and I've been curious what are the capabilities of other teams robots at this time.

[JesseK]

By "Hoarding" balls, do you mean 20+, 30+, etc?

[jamie_1930]

Originally Posted by JesseK

By "Hoarding" balls, do you mean 20+, 30+, etc?

I mean that you have the capability to gather balls and hold them to prevent other from scoring it's broader than the other fields and I see it that if you have no way of storing the balls but in stead shoot and throw as soon as you make contact with the ball then it doesn't count pretty much all others apply.

[Chris is me]

I think you're missing a few functions the most competitive FTC robots will have, honestly. And I doubt anyone who has them will say anything about them until they win their events with them.

[Rick TYler]

Originally Posted by Chris is me

I think you're missing a few functions the most competitive FTC robots will have, honestly. And I doubt anyone who has them will say anything about them until they win their events with them.

Ah, heck, I'll say it.

Look at video of Aim High FRC robots for inspiration. With one exception (an amazing quick-fire catapult) they all used some version of a baseball pitching machine. Both single-wheel and double-wheel versions worked well. If you want to shoot, it will look like a pitching machine, and mostly the wheels will be top-mounted to minimize blocking.

Picking up balls is a pretty easy exercise, as is shooting them. Tying the shooter together with a ball store and picker is the engineering challenge. I predict that a lot of teams will fail at the center mechanism.

Check out the Robowranglers (FRC 148) Lunacy robot for some great inspiration on how to solve this problem. Add a baseball pitching machine to the top and Bob's your uncle.

You can vary the distance and height of your shooter by either varying the shooter wheel speed or the point of release. In 2006, I preferred the second method, but there's not much question that either can work.

A mechanism or strategy that can turn the goal is a valuable thing. I'm not suggesting any particular mechanism, but I have noticed that the penalty for moving the goal in autonomous is pretty small in comparison to the points you can avoid <deleted>. In the middle of writing this paragraph it occurred to me that the GDC is a pretty bright bunch and that they might have noticed that a 5-point penalty isn't enough to dissuade this strategy. In the latest version of the manual the penalty for SG7 has been increased to 40 points. Take note! A relatively minor offense is now the death penalty.

Most FTC robots I saw last year were poorly controlled, and I don't think this was bad driving. The drive motors are so powerful in relation to the size of the field that teams built fast robots that were hard to drive. I expect the best robots this year to have great smoothed-and-scaled drive code. With another cluttered field this year with four relatively large robots on it maneuverability will be at a premium, so I expect to see new-generation driving code to be combined with holonomic drive trains in many of the best robots. Our new FTC robot uses a holo drive and it's a revelation compared to last year's machine.

The blocking rules may engender some strategies which include "drawing fouls" by shooting balls in such a way as to lead referees to believe that another robot was blocking. The reliance of the rules on having referees divine the intent of drivers will lead to mind games where teams try to convince the referees that the other alliance intended to commit a foul where none was meant. I'm not a big fan of rules that require referees to figure out what teams intended, but it does open up a rich avenue for strategy boffins.

Of course, you need a good strategy for acquiring balls. Sucking them up as they come down the ramp seems like a no-brainer to me.

The ability to hold zillions of balls looks good, as is the ability to hit the off-field goals from more than minimum distance. Shooting while moving would also be valuable, but I doubt we'll see much of that.

Just like last year, winning auto will mean winning the match nearly all the time. The auto bonus is so incredibly huge in the last two FTC games that if you can't score in auto you might as well not play. Conversely, a terrific auto defensive strategy might make you a popular alliance partner. A defensive robot won't get you to Championships, though. Top scoring robots will end up top-ranked after qualifying, and while a great defensive robot will be a valuable alliance partner, it will only get ranked high in qualifying if it is paired with a great scorer. Since only alliance captains go to Champs, I wouldn't waste time building a defensive robot. A great defensive team would be a good partner for the best scorer, but won't go to Worlds.

Got to go back to work, but that's my 10-minute analysis.

[Chris is me]

After quickly realizing I don't have a team, might as well add some thoughts to Rick's excellent analysis on what I see as stuff that seems particularly important. Disclaimer that I could be very wrong.

Originally Posted by Rick TYler

You can vary the distance and height of your shooter by either varying the shooter wheel speed or the point of release. In 2006, I preferred the second method, but there's not much question that either can work.

Though I don't know how you can build a sliding release point in FTC, I do know sliding release points (especially in Lunacy) resulted in much faster firing, to the point that notable teams this year that relied on wheel speed would abandon ranged firing altogether in favor of building a hood permanently tilted.

A mechanism or strategy that can turn the goal is a valuable thing. I'm not suggesting any particular mechanism, but I have noticed that the penalty for moving the goal in autonomous is pretty small in comparison to the points you can avoid <deleted>. In the middle of writing this paragraph it occurred to me that the GDC is a pretty bright bunch and that they might have noticed that a 5-point penalty isn't enough to dissuade this strategy. In the latest version of the manual the penalty for SG7 has been increased to 40 points. Take note! A relatively minor offense is now the death penalty.

40 points, depending on the strategy, might not be the end of the world. I'm not saying it's anything you should bank on but it certainly will happen in at least a few matches with the specific intent of messing up autonomy. Considering auto points are worth double (basically), even if they pick up all the balls they miss or they don't fire if the goal gets messed up, you could potentially deny more points than would be scored in very extreme circumstances.

Most FTC robots I saw last year were poorly controlled, and I don't think this was bad driving. The drive motors are so powerful in relation to the size of the field that teams built fast robots that were hard to drive. I expect the best robots this year to have great smoothed-and-scaled drive code. With another cluttered field this year with four relatively large robots on it maneuverability will be at a premium, so I expect to see new-generation driving code to be combined with holonomic drive trains in many of the best robots. Our new FTC robot uses a holo drive and it's a revelation compared to last year's machine.

If I understand what I saw correctly a lot of this was trouble turning, which omni wheels will help with. Holonomic drives will excel at chasing the 5 point goal in particular. If I were building an FTC robot I would go 6 wheel for other reasons though.

I would not be surprised to see a swerve chassis of some sort this year.

Of course, you need a good strategy for acquiring balls. Sucking them up as they come down the ramp seems like a no-brainer to me.

This is what I've been trying to push on anyone who will listen. Almost every team I've talked to just wants to deploy the ramp then use intake to get them from there, and that's giving your opponent free points. Controlling the gates with a drop down hopper or just some "wings" or something is more important in my mind than human loading in Lunacy or Aim High. I wonder if you could get away with no floor intake at all if you really had to.

The ability to hold zillions of balls looks good, as is the ability to hit the off-field goals from more than minimum distance. Shooting while moving would also be valuable, but I doubt we'll see much of that.

I honestly think the 10 point goal is the most important part of the game. Scoring in it is just as valuable as scoring in autonomous, but easier as you have drivers to control it. The goals are also wider so you can shoot 2 or 3 balls wide into the goal, unloading at ridiculous speeds. Just one corner worth of balls plus a doubler and your preload and you break 400 points. 40 point penalty doesn't look so bad now.

[JesseK]

Think about what it takes to prevent auton. scoring by spinning the bar:

• A bot must get over the center wall consistently or have a very finely designed mechanism to move the bar from outside the wall without breaching the 10" boundaries
• The bot's alliance partner also foregoes the ability to score in the center goal, at least at the regional level, should spinning the bar be successful
• If the bot does make it into the middle, it risks the opposition getting into position to take your ball chute as soon as teleop starts, therefore netting them an advantageous 3 chutes to 1 chute
• Even if the bot does not go into the middle yet spins the bar, and the opposition's bots are smart enough to ignore the auton. shot if it's not where it's expected to be, the bot is still on an even field with them -- 2 chutes for it, 2 chutes for the opponents, except now the bot is 40 points down.
• Let's say the bot tries to go into the middle but doesn't make it one match -- it won't incur the 40 points, but essentially it's now a busted autonomous, which leaves the door wide open for the opposition to score as well as getting more ball chutes than you. A little bit of defense in the endgame, and the opponents would win.

It's definitely a niche strategy that could win in the clutch, but overall it would be very very risky to design for without having already seen the game played by actual teams. I'd put it on par with a Goldman Sachs investment strategy: it's great at the moment for you, but bad for everyone, including you, overall in the longer term.

Additionally, playing defense in auto could also lead to a disadvantage at the ball chutes ... though I'd estimate the chances of that are the same chances two robots block each other on the way to getting their far chute in auton. It's just too difficult to tell how it'd play out right now and so it's definitely an option we'll have available.

[Chris is me]

I was personally thinking of getting a close ball chute locked up, then going for the bar (with line tracking or whatever). Partner could secure the far chute. If the alliance only scores their 4-8 preload into the high goal, well, worse things have happened.

Still, this strategy will rarely ever happen. I wouldn't count on it.

[FTC G-FORCE]

[quote=

I honestly think the 10 point goal is the most important part of the game. Scoring in it is just as valuable as scoring in autonomous, but easier as you have drivers to control it. The goals are also wider so you can shoot 2 or 3 balls wide into the goal, unloading at ridiculous speeds. Just one corner worth of balls plus a doubler and your preload and you break 400 points. 40 point penalty doesn't look so bad now.[/QUOTE]

Ok if you go for the spin bar, then by the time you get there the other robot has the 40 points. IF they can get in range and gets them in right away. THEN its like they have 80 points (I don't think its the best way to win) + they only get 40 points if they get all 8 balls in with alto mode UNLESS they get the YELLOW ball in at the end.

[Chris is me]

Originally Posted by FTC G-FORCE

Ok if you go for the spin bar, then by the time you get there the other robot has the 40 points. IF they can get in range and gets them in right away. THEN its like they have 80 points (I don't think its the best way to win) + they only get 40 points if they get all 8 balls in with alto mode UNLESS they get the YELLOW ball in at the end.

The strategy would be tuned to the opponent; if the opponent can shoot all 10 of their balls accurately into the goal within the 5 seconds or so it takes to drive forward over the bump and push the spin bar with a spear or whatever accessory of choice is in use (assuming the robot would hit the goal), then the robot could run a "normal" outside goal grabbing auto.

Remember if a robot makes just as many balls in the 10 point goal as the opponent makes in auto, they've canceled out their advantage. So then if the 10 point shooter ends up grabbing a goal worth of balls and scoring them in the end, they could get as many / more points than a team scoring in the 5 point during auto and the rest in the 10 point goal, assuming same number of obtained balls. Add in doublers and the ten point goal basically has the most potential of all of them. granted, this requires a massive hopper and a very fast shooter, so this strategy only works if you had that...

It's an interesting tradeoff to figure out which goal to aim for (or both, then decide at the event), but I think a properly designed 10 point only robot wouldn't NEED to win auto to do super well. However, the event winners will probably be 5 point auto scorers, 10 point teleop scorers...

[JesseK]

Chris, are you thinking that you're able to score into the outer goals during auto? It's my understanding that no one is allowed to score into the outer goals until the last 30 seconds of teleop.

[Chris is me]

I'm thinking ball for ball, a ball scored in 5 point auto is the same as a ball scored in the 10 point goal.

[FTC G-FORCE]

yes but even if they did'nt get any of the 8 balls in the 5 point goll and they don't get the 40 points and you move the spin bar then your partner and you need to save your 8 balls then when you spin it then you have -40 and they have 0 and the 8 balls are on the floor i can't see how that will help (but its not my team so do whatever you fell like) + by doing that in auto mode i think the judgs won't like that and if other teams see you doing that they might not won't you as a partner.

[jamie_1930]

Originally Posted by Chris is me

After quickly realizing I don't have a team, might as well add some thoughts to Rick's excellent analysis on what I see as stuff that seems particularly important. Disclaimer that I could be very wrong.



Though I don't know how you can build a sliding release point in FTC, I do know sliding release points (especially in Lunacy) resulted in much faster firing, to the point that notable teams this year that relied on wheel speed would abandon ranged firing altogether in favor of building a hood permanently tilted.



40 points, depending on the strategy, might not be the end of the world. I'm not saying it's anything you should bank on but it certainly will happen in at least a few matches with the specific intent of messing up autonomy. Considering auto points are worth double (basically), even if they pick up all the balls they miss or they don't fire if the goal gets messed up, you could potentially deny more points than would be scored in very extreme circumstances.



If I understand what I saw correctly a lot of this was trouble turning, which omni wheels will help with. Holonomic drives will excel at chasing the 5 point goal in particular. If I were building an FTC robot I would go 6 wheel for other reasons though.

I would not be surprised to see a swerve chassis of some sort this year.



This is what I've been trying to push on anyone who will listen. Almost every team I've talked to just wants to deploy the ramp then use intake to get them from there, and that's giving your opponent free points. Controlling the gates with a drop down hopper or just some "wings" or something is more important in my mind than human loading in Lunacy or Aim High. I wonder if you could get away with no floor intake at all if you really had to.



I honestly think the 10 point goal is the most important part of the game. Scoring in it is just as valuable as scoring in autonomous, but easier as you have drivers to control it. The goals are also wider so you can shoot 2 or 3 balls wide into the goal, unloading at ridiculous speeds. Just one corner worth of balls plus a doubler and your preload and you break 400 points. 40 point penalty doesn't look so bad now.

I wouldn't count on anyone trying to spin the bar in autonomous because by doing so they're saying that at least 8 balls will be scored in the high goal and I think that's very unlikely to happen.

[FTC G-FORCE]

o and every time you hit the spin bar you - 40 points so if you bump it 2 times you get -80 points