Re: Infared in Next Year's Game
 

Well geez, arbitrary rules placed on problems is a staple of mathematics! (I say this just getting home from a Linear Algebra class ... so forgive me if my mind is on math, but I just got through solving one such arbitrary problem.)

And, if you think about it, engineering too ... the rules aren't too "arbitrary," though, so long as they make you learn something new and delve into unfamiliar territory (a good thing). E.g., we're limited to 130 lbs., we must only use certain motors and in certain quantities ... etc. These are called constraints -- and mathematics, as well as engineering is full of them! Using IR sensors could very well just be another constraint.

Re: Infared in Next Year's Game
 

Fair enough. I dont think there should be no constraints, but the other constraints are a little more realistic in my opinion. They seem more like real ones, and they often have practical reasons in our case, they dont just mimic practical constraints :) Like the weight limit is what two students can probably carry and the size limit is less than a wide door. I guess its mostly my opinion as i can't put it into words, but i think that forcing the use of a sensor is unrealistic.

Re: Infared in Next Year's Game
 

they force us to use:
the RC system
the battery
the main breaker
the smaller breakers
the C programming language
the OI
the same motors
the RF link
......

if they somehow forced us to use the IR sensors and beacons next year to play the game, its nothing they havent been doing for years.

Re: Infared in Next Year's Game
 

pracitcal reasons for everything on the list, safety, money, fairness.......... maybe there is one you can use to argue for the IR sensors, fairness....

Re: Infared in Next Year's Game
 

yes, the object of the game

this year the object of the game was to hit or pickup small balls, place large balls ,and hang on a bar

they could easily make the object of the game be to find an object that is emitting an IR signal - maybe only a few out of many emit the IR

its would be an interesting challenge because IR is invisable to humans, so you could not fall back on having your driver make the selection based on what he sees on the field

I think it would be fun.

Re: Infared in Next Year's Game
 

I agree with you completely. Though it would make the game a little more interesting to have an element of IR or something of the sort required, it's just not in the FIRST spirit to make it mandatory that you do things a certain way, is it? I=Inspiration (or improvisation, depending on who you ask), and you don't get inspiration from someone telling you what you must use on your creation. Right?

Re: Infared in Next Year's Game
 

I think it would add a completely new dimension to the game.

I would love to have objects that are IR-tagged and invisible to the driver between the right objects and wrong objects. The only problem I see is it could make the game extremely boring to an audience. Did they score or not? So I don't see that happening unless they have some way for the audience to know the right ones (think about tv game shows where the tv audience can see the price and they are yelling at the moron on tv who isn't even close.....I watch too much Price is Right..lol.) Now that could be cool but I don't see how it would work.

The one thing FIRST would have to do is offer multiple scoring options (non-ir) or have code and general instructions to get IR working and have it finished before kickoff so their IRbots can battle it out to show us the game. No more volunteers on bikes or ditching non working robots to have the humans play. If FIRST can't get it right how could they expect the teams to.

Re: Infared in Next Year's Game
 

they could play in total darkness, and give everyone in the audience night vision goggles :^)

Re: Infared in Next Year's Game
 

Yes but I don't know if I trust the live audience. There would be a lot of signaling to teams. Or at least yelling out, "yes" "no" "left" "right". I guess it could be like the price is right except the audience knows the answer.

Re: Infared in Next Year's Game
 

I guarentee you this will not be the case since it is fairly easy to see infared using a camera. They would have to randomly change the feild every single time or else there is no challenge.

Re: Infared in Next Year's Game
 

and you think the driver would be able to hear any of that?!

have you ever actully BEEN to a regional? :^)

Re: Infared in Next Year's Game
 

I dont understand your reasoning? the drivers would not be allowed to use a camera to scan the field, or anything else

Re: Infared in Next Year's Game
 

They wouldn't hear. But people would still try. And after the match was finished and the driver made a mistake they would come up and say "What were you doing, didn't you hear me."

I was more referring to signals such as people with huge signs. You could make a really cool LED one that you could push a button to light up the words. Now that would be awesome. If you ever saw 34's light up LED rockets you could see what I mean.

Re: Infared in Next Year's Game
 

or you could hook an IR sensor to a servo on your bot, and have the bot scan the field with about 3 lines of code - much easier than building an LED alphanumeric display panel :^)

Re: Infared in Next Year's Game
 

Boy oh boy, at the regionals I attended, a LOT of people were VERY ticked off about the way IR was implemented this year, and were simply reprogramming their robots to run without their IR sensors...

The biggest challenge I see is the mere fact that we have two competing forces at play:
1) The desire for the arena to "look flashy" for the TV CAMERAS, and
2) The arena be "Robot IR Vision Friendly".

The choice of shiny diamond plate on everything was nice for the cameras, but caused a LOT of teams' hardware to see IR beacons everywhere. They were grumbling about the Kickoff's "solemn promise that we'd have reflections dampened", and people swearing they'd NEVER be suckered into working on something like this again if there was ANY simpler way to do it. <sigh>

Yes, some had designs that via one method or another you could ignore or filter out reflections. Most simply preaimed their scanners toward the primary emitter of interest at round start, or ignored things until their robot was CLOSE via ded reckoning. But what's the fun in that? You shouldn't HAVE to do such things.

Note also that you still didn't NEED IR this year, nor line following. In fact, many people using ANY "binary" sensor based steering system (e.g. primitive narrow focus line followers that said "go left, go right") quickly found themselves driving very slowly to hold the line or path without oscillating, and were beaten to the tee by a good ol' ded reckoner almost every time... NOT the kind of negative feedback you want to give people TRYING to use the advanced method!

I feel that as long as FIRST definitively specs object locations on the field, autonomous time is so short, and the given sensor/cpu combo is crude, there's no NEED to risk ANY sensor based approach beyond ded reckoning. You're just "asking to get your bot kicked"... ;) And honestly, that's too bad. I'd LOVE to see some serious sensor based activity in this game! (Don't mind me... Among other things, by trade I'm also an Instrumentation Engineer... :D )

IMO, What FIRST needs to do is spec a "random position object" or otherwise TAGGED item in their game specs. IMHO, it shouldn't be the WHOLE game, but a BONUS.

My advice to make IR (or optical systems) more important:

1) KILL the IR reflections on the field! Either get rid of shiny diamond plate, locate the beacons where the robot CAN'T see it, or find a method, coating, or material that makes things nice and shiny in visible light for the cameras, but "flat/absorbing" in IR. That would be VERY helpful.

2) Consider placing redundant beacons or corner cube retroreflectors around (or on the corners of) the field, as navigational "landmarks". I say redundant, because robots and objects will always occlude SOME beacons. Color filtered retroreflectors could make great landmarks, as long as you provide emitters, detectors, and/or the corresponding color filters (or software) in the kit to place in front of the optical detectors to differentiate them.

3) Include a marked, yet randomized object as a field device. Give extra credit to those that ID, find, or manipulate it CORRECTLY.

One example: "Find The Beacon(s)". When you see an object in front of you, do something like push a bar on it to raise a flag for points, shine YOUR color of light over its well to change ITS color so you "CLAIM IT", or move it into the CORRECT basket. Deduct a LOT for "guessing wrong". Leave traps so blind random attempts to tag EVERYTHING takes points AWAY from your alliance.

Another example from this game might be to have a polygonal sided rolling goal with multiple vertical posts around its perimeter. One of them, and its opposite mate, are marked with a beacon (or is shiny, vs flat black). The goal is again placed with one edge against some stop, in a randomized fashion. This gives each side a shot at a "valuable" post. Grab and manipulate it by the "valuable" post and you get more points than if you grab any other. This also allows SOME chance that by sheer dumb luck you can gain points if your scanner fails, but rewards those that "do the work" more consistently. :D (Field attendants roll a multisided die, and set the orientation AFTER you are ready, robot turned on but disabled, and all players waiting for round start in the player's box, with their hands off their controls. This prevents "signaling" the robot's auto mode code with "the right answer" of several options via a button press.)

A third example: Instead of COLLECTING, for once make a "DISPENSER". (Perhaps even be given a preloaded dispenser to snap onto your robot??) You now run around flagging beacon marked randomizable objects by attaching them to the object (maybe snap a detachable carabiner onto a ring or bar by the beacon?), or running back and forth between two rows of baskets, placing YOUR color of ball into the basket whose beacon is turned on THIS time.

4) Make a SLOW, RANDOMLY MOVING object, that's beacon marked, such as a slow turtle, that's patrolling a short line, circle, or figure 8 of tape. Find it, touch it, grab it, or tag it with a flag, and you get points. (A favorite object of mine is a "Santa Ball", those crazy running motorized balls that go careening off of walls... Add beacons to them and let a few loose as the Bonus Balls!)

5) Make a spinable turntable that you have to interact with (like in Toroid Terror") or a vertical "carnival wheel" whose ORIENTATION is indicated via a beacon or a corner cube on its perimeter. Lock it so it can't simply be spun by the robots. That orientation beacon/marker tells you where to go to access it.


There are a LOT of possible variations here. The common thread is always use the marker to indicate something on the field that is otherwise UNKNOWN, at least at the time autonomous mode is started.

- Keith