Hanging vs. Manipulating 2x balls [Archive]

me9342

11-01-2004, 18:05

Say you can pick one of the above - your robot will be able to either hang from the bar or be able to place the 2x multipliers on a container. (As much as I'd like both, I think that it would be both impracticle given the time limit, and impossible due to the weight constaints.) Which would you pick? I myself would go with a 2x ball manipulator, as it has the higher potentional for total points (if you have 11 balls or more in one container, you would get more points for having it capped than you would for leaving it uncapped and hanging). I realize that if you dont get many balls, or have a really bad human player, you won't get many points for capping, so I can see some reason in just taking a flat 50. I wanted the opinions of others, though.

animater31405

11-01-2004, 18:07

Nothing's impossible. If you have arm that has the ability grab ahold of the high bar using the SAME arm pieces to grab the 2x ball it is extremely possible to do both. That is the way we are talking about going.

me9342

11-01-2004, 18:14

ok, its not impossible, (we came up with the beginning of an idea that could do both), but for this purpose, assume you could only do one.

animater31405

11-01-2004, 18:16

I would probably go for hanging because there's a better chance that someone else will want to go for the ball and if they don't they can get your alliance to score 100 easy points.

kevin.li.rit

11-01-2004, 18:20

I would probably go for hanging because there's a better chance that someone else will want to go for the ball and if they don't they can get your alliance to score 100 easy points.

Wouldn't they only be able to score 50 points?

andy

11-01-2004, 18:28

Mercutio

11-01-2004, 18:39

The thing about the multiplier balls is that it seems pretty easy to push them off, while if you got a good grip on the bar you would be practically immovable. You'd get knocked around a lot, of course, but you'd stay on. :D

If you could make a multiplier manipulator that could grab the bar too, like animater31405 said, you might do *very* well. I think the key in this competition will be doing a lot of things with the same part. If you try to make a separate mechanism for everything, you'll end up hitting the size/weight limit about a third of the way in. If you try to specialize, you'll come up against an opponent who can defend against your particular strength.

~Aaron

P.S. I know someone will show up at regionals with a robot that whacks hanging bots with a baseball bat. LOL, piñata time! :p Send pictures!

yangotang

11-01-2004, 18:42

Mercutio

11-01-2004, 18:46

i'm convinced that a 2x ball remover is the easiest strategy.
No argument there. But removing the things is a *lot* eaiser than putting them back on...

i feel that creating a method to get up to the platform below the bar is tough
If we actually get ours working, I'll show you how it's done. ;)

~Aaron

westfalia

11-01-2004, 21:27

the problem with the 2x ball is that it can't be your only objective. I agree that you've got to use one thing on the robot for many functions. but it is hard enough deciding upon a good idea to get to the top of the bar!

Aignam

11-01-2004, 21:40

By far not impossible. A slightly modified version of Team 25's 2000 robot, Cradle Robber, could more than likely do that.

KenWittlief

11-01-2004, 21:52

Double your pleasure

Double your fun!

Greg Powers

12-01-2004, 01:42

By far not impossible. A slightly modified version of Team 25's 2000 robot, Cradle Robber, could more than likely do that.

As a member previously on Team 232 I actually still have the video footage of that bot in action in the 2000 finals and we are seriously considering doing just that. (moding that design)
It would be difficult, but fairly ideal for this years game if executed well.

Aignam

12-01-2004, 06:00

As a member previously on Team 232 I actually still have the video footage of that bot in action in the 2000 finals and we are seriously considering doing just that. (moding that design)
It would be difficult, but fairly ideal for this years game if executed well.
It is a good design, and might be quite applicable for this year's game. The only problem that I see is that in 2000, we cherry-picked balls from our opponent's goal and put them in our goal or moved balls from our goal to theirs, controlling the score. In this game, that would be quite illegal. Beyond controlling the 2x Ball and hanging on the bar at the end of the match, the robot would be pretty useless. Unless, of course, you build it with a robust drivetrain, such that it could plow balls into the human player corral with ease. But is it safe to build a robot that can do more or less everything?

Mikey229

12-01-2004, 08:52

I think doing BOTH is a very good idea as yes you can do them both with the SAME arm...

Dr.Bot

12-01-2004, 10:16

If you want to control the game, get one or two of the Big Balls, Hang High, and then decide if you need to score them for both you or the opposing alliance (Stationary goals). You can then easily add points to the opposition. (You also might be able to uncap the oppositions stationary goal.)

This was the winning strategy in 2000 by team 255. Get the balls, Hang on the Bar, and then score them.

Is this mechanically feasible - I don't know. A team that does this needs to rely on its partner to score the small balls. A perfect robot would be able to hang, score big balls, herd little balls - and this is pretty challenging.
Just buildng a robot that can climb the big steps and hang is a challenge.

A smart development plan would be to design by objective:

1. Build moving platform cable of herding, releasing
2. Add Arm for big ball manipulation
3. Add stair climber
4. Add hang ability

6 weeks, 5, 4, 3, 2, 1 ship.......

animater31405

12-01-2004, 12:48

Don't forget that hanging is really two challenges at once:
1)being able to get up to the bar
2) latching onto the bar

While these two challenges may seem easy to you, i feel that creating a method to get up to the platform below the bar is tough; i'm convinced that a 2x ball remover is the easiest strategy.
alright i also think that if you create a good arm you could actually pull yourself over the platform from the ground floor and lift yourself up.

KenWittlief

12-01-2004, 12:56

I think doing BOTH is a very good idea as yes you can do them both with the SAME arm... it will be interesting to see how many teams try this

because having one mechanism that performs two functions violates one of the Golden Rules Of Engineering (GROE)

"a subsytem should have ONE function, and be optimized to perform that function"

when you increase the number of functions a subsystem has (N), you increase its complexity N^N

so the complexity of a subsystem with one function is 1

the complexity of a subsystem with two (different) funcitons is 2^2 = 4

the complexity of a subsystem with 3 functions is 9...

it gets out of control REALLY quick!

animater31405

12-01-2004, 13:01

it will be interesting to see how many teams try this

Our team has already drawn out a device that can do both the ball and the arm.

Dr.Bot

12-01-2004, 13:09

The mechanism that extends the arm, if powerful enough, whe reversed can lift the robot (1 motor, two functions) Once you latch the bar, you only have to lift a little to hang. I don't think a lot of teams will try to win by
messing with robots trying to hang. You want to score points, not try to prevent your opponents from scoring points.

A simple way to do this is a chain drive elevator using the door motor.

--- 2000 robot using chain elevator to lift scoring mechanism, hanging from bar.http://robotics.nasa.gov/foothill/pictures/7a.jpg

lips

12-01-2004, 13:09

there is also the possibility of creating a small hanging device (schematics forthcoming, plan already designed) and then having a large manipulator arm. this would allow a team to both hang, and while hanging knock off a ball from one (or either) side if stragegists deemed necessary.

Just a thought

KenWittlief

12-01-2004, 13:11

good start - now draw out what it takes to do each one separately

if you need incentive - if your 2X ball arm is damaged while playing, you have lost your 50 point hanging function too

thats called a single point failure - if one device does both functions, then one single failure in the arm cost you BOTH functionalities (another reason for the Golden Rule)

think about what it will take to do each function separately. You will find the complexity of the two devices is much lower than the single dual purpose one.

ngreen

12-01-2004, 13:37

Otherwise, build robust and prepare to fix parts on that major component.

IrisLab

12-01-2004, 17:28

it will be interesting to see how many teams try this

because having one mechanism that performs two functions violates one of the Golden Rules Of Engineering (GROE)

"a subsytem should have ONE function, and be optimized to perform that function"

...

The key word is FUNCTION and not necessarily ONE. A ramp in the sidewalk has one function, but it has many uses. The function is---simply---a ramp, but folks with wheelchairs use it, as do Coke delivery men with dollys, as do Segway's that must cross the road.

My intent here is that if you define your one function properly, it can have multiple applications. Function does not necessarily have to be defined as hanging or ball removal. That could be the application.

To continue with the analogy, if the function of a ramp in the side of the road was to allow Coke delivery men to move their cokes from a truck to the machines in a building, the ramp would only need to be wide enough for a typical dolly. Unfortunately, a wheel chair or Segway wouldn't fit up such a narrow ramp.

I know this analogy is simplicistic, but I hope it highlights my point.

animater31405

12-01-2004, 17:35

if you define your one function properly, it can have multiple applications.

That's my point exaclty!

Lil' Lavery

12-01-2004, 17:42

I would chose the 2x mult because it can do a lot more than just score points for your alliance. My team ran some "stubot"(student robot) games and discovered how efectivley it could keep the opponent from scoring in thier goals because their is no way in them. By capping the goals when they have no balls in them, they lose a significant amount of time to score. If you can defend the capped goals then remove the caps so you can use them near the end of the game, your opponent will have little to no points from the balls. Also the bar this year is far less accessable than before because of the platform and the narrow room beside the goals. Thus, you dont only need the abbility to lift and hang yourself from it, but either a narrow robot that can maneuver well and climb the stair, or one that can get over the larger ledge to the section of the platform the bar sits on and still be able to get under the bar. The bar is far more trouble than its worth, because 5 5-point balls and a cap can match its points. And needless to say the goals can hold many more than 5 balls, and their are the 4 10-pointers as well. I beleive that the bar takes up to much game and build season time for what it is worth and that the 2x balls are possibly the most pivotal objects in the game.

dragonpaulz

12-01-2004, 17:45

Do the 2X balls have to be inflated? Maybe you can pop the ball and put them in the goal.

Lil' Lavery

12-01-2004, 17:55

Do the 2X balls have to be inflated? Maybe you can pop the ball and put them in the goal.

Your not allowed to intentionally damage the field, so yes they have to be inflated.

dragonpaulz

12-01-2004, 17:55

OK, thanks.

KenWittlief

12-01-2004, 18:25

I dont see how you can define the function of controlling the large 2X balls:

-acquiring ~2' dia, 3 lb ball from a goal, from the center platform, or from the floor
-placing them on the 4 foot or 6/8 foot goals
- removing them from the goals
-holding them securly during the match to retain possesion

the same as climbing the bar:

-reaching straight up 10 feet
-latching onto a 3" dia pipe
-lifting a 130 lb machine straight up

this is what I was talking about with complexity going 4X when you try to combine the two - the degrees of motion required are different - the end effector is different - the amount of weight to be moved is different - the direction of motion is different

combining the two will make the resulting mechanism much more complicated than it needs to be. I can picture a very simple spring loaded mechanism that can be released once to extend up 10', and capture the bar - and pull the bot up - it only needs strength in the downward direction

but to move the balls you need something like a robot arm - making that strong enought to lift the bot with instead of only moving a 5 lb ball around is a whole nuther ball of wax.

and again, if your multi function arm breaks - you lose big time - you lose BOTH functions on your machine in that match - sure you can fix it later, but you cant get those points back.

animater31405

12-01-2004, 18:27

You'll just have to watch team 314 and see!

KenWittlief

12-01-2004, 18:31

be very mindfull of single point failures that will disable your machine, or that will disable more than one function.

Steering is a good example. If you have a two wheel bot, that steers like a tank - then you have two (potentional) drive motors, but if one fails you cannot steer - all you can do is spin in circles - so a single failure cost you both motion AND steering

but if you add a third wheel and a steering mechanism, then if one of your wheel motors fails you can still move AND steer - and there will be almost no load on your steering motor, so the chance of it failing is much lower - but even if it does fail you only loose one function, steering. Your bot will still be able to move forward and back - and maybe with tank control of the two drive motors you could still steer a little.

Redundancy is a powerful design concept.

animater31405

12-01-2004, 18:35

True that. True that. But I'm not worried about it getting damaged because we are not playing Battlebots! :D

Ragin_Kage

12-01-2004, 18:39

True that. True that. But I'm not worried about it getting damaged because we are not playing Battlebots! :D

The keyword was: Robust

There is always the possible shoving match, things do break, thats what these 6 weeks and the pits are for, the intent to damage shouldnt be present but you know things happen, just plan ahead, its not an uncommon site

KenWittlief

12-01-2004, 18:43

maybe damaged is the wrong word to use. Failing is more likely to be a problem. These machines we create are hand built one of a kind prototypes - subsystem failures are common during a match - parts break under strain, fasteners come loose, pnuematics develope leaks, motors burn out, all the noise leaks out of your bearings...

thats why single function subsystems is an Engineering Golden Rule - it was learned the hard way - it was learned at great expense - its tempting to think the Golden Rules dont apply to me for some reason

thats why we have then etched in Gold :c)

animater31405

12-01-2004, 18:43

Trust me! *evil grin* I'm not worried about that!

Jeremy

12-01-2004, 20:37

Well, I think the hanging is more important than the 2X balls. :cool:

kevin.li.rit

12-01-2004, 21:28

Well, I think the hanging is more important than the 2X balls. :cool:

6 starting balls + 2x = 60 points, more than hanging.

Lil' Lavery

12-01-2004, 21:46

Even though we arnt playing battlebots, a ton of damage is still done to robots during the competition. Whether it be a shoving match for a mobile goal, or the bar, or a malufunction in the autonomous code, or falling from the bar, robots can suffer some pretty harsh blows. I have thought of some ways to accomplish both tasks, but i personally dont really want the bar anyway, so ive put more thought into the ball part of it. Simple things, such as a broken braze, could put your robot, or a function of it, such as driving, out of comission. Any1 who was at the VCU(richmond) regional last year, we were the robot that had the crappy swinging arm and didnt move most the competition, and also placed dead last. A braze broke in our custom gearbox, which we only had to braze the part because of a shipping error, and immobolized one side of the robot. Thats an example of how a small thing can cost you big.