Re: You write <G34> and <G35>
 

Most likely, I will be head ref at IRI. As long as I get approval from the IRI leadership committee, I would be willing to run a "test drive" of a yellow & red card system.

But, I am not going to wear one of those goofy soccer ref uniforms. :)

Andy B.

Re: You write <G34> and <G35>
 

I like the yellow and red card idea. It will help teams drive their robots according to the rules.
I don't particularly like the idea of the referees deciding what violates such ambiguous rules as tipping, damage, and agressive. They are great people, but I would rather see some official examples shown to all before the referees rule on a play. The referees are human volunteers (like a lot of us), and only "FIRST officials for a day."

Re: You write <G34> and <G35>
 

Maybe not a whole uniform, but what about the zebra hard hat? :)

Re: You write <G34> and <G35>
 

I dont understand why you feel this way? do you play any sports in school? do you take gym class?

if you dont get between a bot and its goal and it gets into scoring position, then you lost your chance - its too late - why do you think its now ok to push it over and 'take it out' for the rest of the match?

put it in the perspective of other sports. If a batter hits a ball far out into the field, and is rounding 3rd base, do you expect the 3rd baseman to just stand there and watch him score a homerun? why not run into him as he goes past? why not knock him down to give the fielder a chance to throw the ball to homeplate before he gets there?

cause if a player is in 'scoring postion' then they have already played the game better than you. I defy anyone to build a bot that can lift the 2X ball up into the air and cap the stationary goal that is not vunerable to being knocked over, seriously - it cant be done

and the rules state that 'any strategy aimed solely towards tipping... is not allowed'. But Ive seen match after match where this happened, so I dont see how anyone can say, we only intended to push them sideways, we expected there wheels to skid sideways and the bot to NOT fall over, but for some reason it fell over.

I think after this year you are going to see more and more teams who tried to build machines with complex articulated arms for capping, or ball collectors simply come back next year with tanks, armored vehicles that savage anything on the field, cause they tried to play the game to score points this year and ended up watching match after match with their bots knocked flat, and no penalites called.

if thats the way the game is played now in FIRST, why bother trying to cap or score points?

seriously, why bother to build an arm to place the 2X ball if you know you will be toppled over and over if you make it to the elim rounds? Why build a ball collector if someone can fly across the field at full speed and smash your rollers? why build a ball capture machine if someone can stuff their sky hook or claw into it and pull you over

and like I have said several times now, not a single team was DQ anywhere this year for tipping or damaging another teams bot, anywhere.

all those teams who lost out after being knocked over, or who could not continue to play due to severe damage - what have they learned this year?

Re: You write <G34> and <G35>
 

I know that our robot can lift the 2X ball 13 feet in the air and we only fell over 1 time during practice and we popped right back up. We built our robot with the lowest center of gravity that we could. All motors were kept as close to the ground as posible. There is a chance that we could be pushed over but ANY robot can be tipped given the right circumstances. We also built for power and a rough game. When looking at the game we figured that there might be as much pushing as last year. We also built to hang from the outside so that we could keep our center of G low and not have to worry about fighting for a very small place at the top.

In other words we came as prepared as we could. There seemed to be more problems with teams flipping or getting hung up on those stupid purple balls than by other robots. Remember, engineering is not about building an object with the least amout of tolerance. You need to test beyond what you feel is normal. FIRST tests us in so many ways. I personally feel that we should have been allowed more weight because of the complex nature of this years game. FIRST did notgrant this and so our robot needed a diet to lose weight. We had to decide whether we wanted to lose strength or function. The robot ended up losing mostly strength but also some function. We knew what we wanted and hopefully came close.

We need to play the game as it was written just as we have to build to FIRST's specs. To be aggressive and rough is one thing but to deliberately damage another robot is unacceptable.

Re: You write <G34> and <G35>
 

There is a line between defense and agression (as mostly seen from my team's robot 469). For instance: robot A places its arm above robot B's open hopper.

Defensive: rbt-A holds its arm above rbt-B and trys to block balls from falling in.
Agressive: rbt-A bashes its arm continuously into rbt-B's hopper, hopping to damage some system and cause damage.

Most of the current judging is based on how "intentional" an action is. If you look at some of the matches, you will see shoving matches, robots falling over everywhere, etc. The referees had a difficult job this year, especially with all the arms and large box robots that could be prone to tipping. Personally, if FIRST were to remove all defensive maneuvers from all games, then there would be much less variation to the game, and to me, it would be boring to watch
>>"hey joe! look! rbt-C is going to cap the large goal and gain 40 points! shouldn't rbt-D go and block them?"
>>"they can't touch any other robot on the field, so they can't do anything about it."

As for the entanglement issue (see this controversial match), it is not clearly defined. ANY fabric on a robot is prone to entanglement with other teams, so how can this be sorted out? should teams not be allowed to go anywere near fabric-teams because they could get caught? Most teams do not wish to be intentionally stuck on another robot, and unless the action looks clearly intentional the judgement is left to the refs (thus making their jobs more difficult)

overall I believe the current rules are fine for the purposes of defense, but it is difficult to definfe how intentional an action is.

_Alex

Re: You write <G34> and <G35>
 

I think what you're looking for is something like football's pass interference.

<G36> Defensive Interference-
A team may only defend a goal if they are directly in front of the goal or will be interacting with the front of the opposing robot. If the defending robot interacts with the offensive robot in any other way a 10 point penalty will be assesed.

Definitions:
FRONT - The front of the robot is considered the side which is the first part of the robot to arrive at a destination if the robot was driving at something. Crab steering robots FRONTs change dynamicly and will vary in different situations.

Re: You write <G34> and <G35>
 

Ask 469 and 71... They seemed to do pretty good...

Honestly, this is a big debate right now, right after the Championship, among a few people (compared to the total number of FIRST people out there).

I firmly believe that by the time FIRST 2005 comes around, great ideas of innovative mechanisms will take over and we won't have a bunch of tanks driving around...

Re: You write <G34> and <G35>
 

As a preface, our robot played zero defense. I hated teams playing defense against our machine - it was very effective. However, it's part of the game. Getting hit from the side, having arms fight one another during capping, and being rammed on the 6" platform were obvious implications of this years game. I'd like to comment on some of Ken's remarks.

Ken, I'm not a practicing engineer, but I've taken a simple class on static bodies. Any size robot can be tipped over if you push from a height tall enough with enough force to cause the moment force about it's edge to be greater than it's weight.

My point is that, most robots do indeed push from the floor, perhaps 2 - 8 inches above the ground. Robots can be pushed sideways safely, just like you can push a bottle of coke across the table when you push on its base, or it can tipped if you try to put it over at the top.

<R10> Teams are expected to design and build robots to withstand vigorous interaction with other robots.

I believe that being pushed from the side isn't vigorous... it's defense. If you don't want to get tipped over when hit from the side.. you have a few options. You can have wheels with low traction, lower the machines center of gravity, use casters that will cause the robot to pivot instead of turn. I think the major problem that teams had this year is using big wheels. Because they were using bigger wheels, their effect wheel base was smaller, causing them to be much more prone to tipping.

There are engineering choices every year that each team needs to decide for themselves. How to avoiding being tip-prone is one of them.

Isn't it obvious? You don't get 50 points for tipping a robot, you get 50 points for hanging on the bar. You need to build an arm so you can do so.

Hopefully teams that had problems tipping learned that having a lower center of gravity is critical to ensure that you're stable. I also hope teams have learned that if you're going to have a big bad arm, you better design it such that you can pick yourself up - you look really dopey if you don't.

In conclussion, I think that there are a lot of defensive plays that people can do effectively. I think that repeated ramming isn't effective or gracious, but I think that engaging machine robots when they are prone to weaknesses is fair game, ESPECIALLY in the finals. FIRST is a competition, and tipping, though potentially devestating, has been and always will be part of the game. Well prepared teams should be able sustain vigerous interactions.

Matt

Re: You write <G34> and <G35>
 

Ken, I am not an ME, but I think I can design a robot that is extremely hard to tip. I would start by looking at team 67's design. Their entire base is shaped like a wedge. If someone were to run into them, they would more likely drive up on top of 67, transfering weight to 67 and helping to keep them from tipping. They also had a very low base and a thin arm, which means it would be very hard to to push on anything high enough to counteract their low center of gravity.

(I should say they never tipped over, except when they wanted to at the beginning of the match ;))

Re: You write <G34> and <G35>
 

Joe,
You beat me to it.

John

Re: You write <G34> and <G35>
 

I think what Ken means there is that if you hit the robot high enough and strong enough, it will go over.

Re: You write <G34> and <G35>
 

John, I like where you are going with this. I think that most of us could decide by watching a match that a team was stepping over the line into "agressive" behavior. If most of us can tell the difference then we don't need specific rules. A questionable act is warned by a yellow card and that team now knows that at least one other person thinks they have stepped over the line. If they continue then the warning becomes more agressive. Of course, referees need to be knowledgable of the game and it's strategies. I do not envy referees, it must be the hardest job in the organization.

"But, I am not going to wear one of those goofy soccer ref uniforms.

Andy B."
But Andy, the black and white makes you look so official!

Re: You write <G34> and <G35>
 

I think it is possible. Simply make a set of deployable outrigger "lunar landing feet". Given one at each corner of a 5 ft high robot, that is one HECK of a lot of outrigging... :D Add linear extensions and you can span even more. A Backhoe uses this technique. One can lift a LOT, very offset from the base, without fear of tipover. In reality (just like a backhoe), you probably only need one pair of landing feet on the arm end.

But back to biz...

My biggest problems with robot interaction this year focused mostly on things like an arm machine using its big ball claw to deliberately entangle a bin machine's loose netting and yanking in an attempt to drag it away from the drop. Though (to us) an obvious "entanglement move", no DQ's were given at GLR for it, which surprised and bothered us. We had a bin machine that "as delivered" would only be using thin material for the box frame, and a few crosslinked ropes for the bin. VERY vulnerable to COMPLETE destruction from that move. When we designed it, we never dreamed FIRST would allow THAT level of robot interaction.

Luckily, our first regional was Buckeye but we attended GLR and DET just to watch and see how things were run. That revelation prompted us to immediately go out and solve it before OUR regional. We bought some strong ripstop camo netting, and took the raw material with us to cut and overwrap the bin Thursday AM, right after uncrating. It only had tiny holes, which wouldn't allow that move. (FYI, it worked VERY well. Several robots DID try to grab us, but slid off... No one could get a good grip on our now "slippery bin".) :D

Anyway... MY feeling on this whole subject is NOT to bar drivetrain fights and potential tipovers, but instead to simply define some sort of "personal space" AROUND and WITHIN your robot, and simply forbid direct grabbing and object insertion into the opponent's hardware. If violated, it constitutes a "Personal Foul" or "Holding", just like in other sports.

In this situation, direct pushing and pinning on the outside, and possibly even external envelopment dragging would be OK. But instantly banned would be things like: grabbing another robot and dragging it around by sticking in a T-toggle, lifting/wedging it up on a forklift off its drivetrain to carry it out of the way, sticking an appendage into its innards to "grab them by the guts" (and risk "accidentally" pulling out something if THEY resist or even try to retreat), or anything resembling a spearing or martial arts move against your opponent. Think of it as a "personal envelope" around the boundaries of your machine's mechanisms, whose multifaceted virtual planes other machines must NOT deliberately PENETRATE with deployable mechanisms.

BTW - Last year, I envisioned a scenario of one machine's bin knockdown bar/wings being stuck into opposing stacking machine's stacker mechanisms to stop its operation. So I did bring this up with FIRST Q&A, and ASKED them to make some kind of Robotic Personal Space ruling to prevent that. To my surprise they REFUSED to even CONSIDER barring any such grabbing NOR insertion interaction. (OK then, we'll "keep that in mind"...) <sigh>

IMHO, we REALLY need some better guidelines here. My fear is that mech entanglement severity will continue to increase until a better definition emerges for "appropriate" vs "inappropriate" interactions. As it stands now, a team's primary defense may be that any part of another's machine stuck into theirs is fair game for an immediate drivetrain "spin breakage/amputation". I would have NO problem telling MY drivers that if someone has the gall to stick something into our robot chassis, to simply "turn, and drive away with it"... (Souvenirs, anyone???) ;)

I just hope it won't take a few robots torn apart by others before they'll consider instituting a Robotic Personal Space definition of some kind.

[Edit]
Now don't get me wrong. When another robot is about to score, I do feel that "dealignment" moves, and pushing/pulling them are appropriate with the current rule set. They probably should still be allowed in some fashion. Fighting over game resources and positioning are to be expected. I just wish something clearly defined as to HOW you're allowed to engage or grip another, to protect the pull-ee's HARDWARE from damage when it is done ("which wrestling holds" ARE legal).

Basically though, I wish better guidelines that clearly and unambiguously forbid INSERTION and ENTANGLEMENT moves which could interfere with a bot's INTERNAL operations, or rip out hardware. If we don't define that, we may start to see arms with fingers on the end which can simply be inserted into other robots to "toss a wrench" into moving mechanisms as a defense...

If clamping and/or siezing deployable things for dealignment and towing purposes IS to be considered "fair game", IMO the trick will be either in clearly and properly defining what is grabbable and what isn't to limit damage (or defining clear point penalties when damage occurs, whether deliberate or accidental in such a way that still discourages "lizard tail dropoff" style defense). The problem with such a rule is that our bots aren't of a "single species" formula with known structural specs, where you can just clearly say "you can grab XXX but not YYY because it's too fragile"...

That's where I feel a generic Robotic Personal Space definition and rule could be useful. IMO "respecting another's body safety" is the biggest defining difference between playing a game, and brawling. All other sports rule sets are designed to protect the participants' bodies from harm by defining safe grip/push interactions and forbidding harmful ones. I feel we need something similar.
[/Edit]

What do you think? Is a Robotic Personal Space Violation rule appropriate, and/or needed? Would this help solve an "overagression problem"? If so, how would it best be phrased, where are the limits (e.g. what kind of grabbing SHOULD be allowed), and what should be the penalty?

- Keith

Re: You write <G34> and <G35>
 

First of all, I'm an engineer, not a high school student who takes gym class. Second, yes, I played many sports while in school but I don't feel I have to compare this robotics organization to any sport.

Where on earth do I say in my post that I think it's ok to push a robot over and "take it out" for the rest of the match? Because nowhere does it even imply that!

Why can't another robot push us out of the way? You can't honestly expect the only defense of an opponent to be physically sitting in front of my goal. If that's the case, then you would disallow robots on the platform to stick an arm thru the goal in order to prevent my 2x? Technically they're not "in between me and my goal". I think it's ridiculous to say you shouldn't be able to push someone from the side, or front, or back, or other side.
I think that if I built a robot that is top heavy or know that there's even the slightest possibility of tipping if someone pushes on me, I should be expecting it.

No they haven't. Why can't it mean that I have a good offensive robot that is actually playing the game elsewhere, instead of parking itself in front of a goal all day long waiting for you to come along, just so I can "be in between you and the goal". That's silly.

Well by the same token, how in the heck can you say they "intended" to tip you over? Yes, the rule states any strategy aimed soley towards tipping. Pushing is a strategy aimed soley at tipping a robot? Hasn't your team ever pushed another robot? Was your intention to tip them over? Doubt it.
If you look at the matches of 45 and 71, like I previously used as examples, you can see that they've gotten pushed around from their sides without tipping over. So I don't see the harm in that, why should it be disallowed? While I can't speak for 71, 45 built their robot to have the least liklihood of tipping. So yes - Wheels can in fact skid sideways without causing a robot to tip over. It's pretty simple. I would think you'd look at your robot during design and say hmm.. if someone plays defense on me, how easily will I tip? Or, where will I get damaged and how can I improve that?

Eh, I don't think so. I guess you could look at that way if you're negative. But most teams that I saw had an offensive mechanism to help them score points and to play their own offense. If they happen to go against a much better offensive robot, then of course they need some sort of defense. Unless FIRST is invaded with dirty game-playing people, which I hope doesn't happen, then we shouldn't have to worry about the majority of people turning into pure defense robots like you predict. If you watch a lot of matches, I think you'd see that the most effective robot was one that could play great offense, but knew how to avoid/defeat the defense and could play it's own if need be. So I wouldn't expect everything to take a 180 just because some good defense won the Championships.

Good point. Why would anyone bother to build a non-robust robot that can be easily toppled, or easily smashed, or easily pulled around the field? You didn't make any point there. I mean, are you saying that it's not possible to build an arm for a 2x that won't get you tipped? Is it impossible to protect your rollers from getting smashed? Is it impossible to build a drivetrain that prevents you from being pulled around like that? If you said yes to any of those, you haven't see this years robots.

Maybe there weren't many DQ's or calls on certain rules, but MAYBE it's because the refs deemed it "not intentional, not malicious" and within the rules!!! The refs do what they can according to the rules. There's other debates going on about that, how to improve the rules to satisfy more people, but the bottom line is, creating all these conditional defense rules may not be the answer, and certainly not basing them all on what various sports do.
Sure there are some similarities, but I guess I just don't believe you can apply most sports rules that deal with Humans to robots. For one thing, Humans can control their physical actions better than they can control a robots actions. If a human goes running into another human with full body contact, my feeling is that it's easier to call that as an intentional body slam, then you could call a robot doing something similar. I have a pretty good idea that a person will fall over when I run into them with my whole body, especially in plane sight. I don't have a pretty good idea if someone's robot is gonna fall over if I hit them on one side versus a different side, and that's even assuming full view. Maybe I don't have a full view now, how the heck can you penalize me for intentional tipping when I can't see through 2 full goals, pvc, other robots, and a platform? I can't even see which direction a robot is facing. So now are you going to say that I just shouldn't be on the opponent's side at all if I don't have a good view of them? And should I just not shoot balls into the goal so that I can be able to view them?

Common sense comes into play for everyone when judging these conditions. You know darn well when you do something maliciously intentionally, so don't do it. The actions that I've seen debated already don't seem to have been intentional. The only people that know for sure are those people.

It seems so cynical and childish to basically say "why should I bother building a robot at all, that can play offensive aspects of the game, if someone can play defense on me".

How 'bout - lower the center of gravity, don't build top heavy or too tall without precautions, protect your robot features, learn to fix it..... They may have learned those things. Isn't this whole organization partially about learning and improving?