Never Do This, and Other Good Ideas Gone Bad

[Mike Schreiber]

I know my former team, 27, in both '06 and '09 refused to do an Archimedes screw after what happened to them in '02. I don't think they'll ever return to that idea. It does have great potential though, it's made it to Einstein, it's all about implementation I guess.

[demosthenes2k8]

I agree, it's more the implementation than the idea. Several years before I joined the team, they created a robot that remains infamous to this day. Its name was Scorpio. In 2009, several ideas were thrown out because they had failed on Scorpio. In competition, though, we saw them done well. The implementation was flawed, not the concept.

[Dkt01]

Our team will not likely use vacuums, or conveyor belts due to flawed execution and/or excess time we spent fixing parts. We will also never make a robot with less than an inch of ground clearance again (got stuck on the field repeatedly last year).
I agree that almost anything can be successful if done correctly, some ideas are just more difficult to implement than others.

[Jonathan Norris]

Cantilever shafts, try and avoid. Especially over 1.5"-2", try and support both ends of shafts.

[Jon Stratis]

Originally Posted by Jared341

Relying on gravity to do anything other than to give your robot weight. If it moves down (an arm, an elevator, a ball in a hopper), make it powered.

I would argue with this one... If designed correctly, you can rely on gravity. The elevator we built in 2008 was powered up, and gravity down, and the only time it failed to work as anticipated was when the motor literally broke off its mountings at a pre-ship event. Other than that, it always went up AND down when we told it to, at a consistent, reliable speed.

[EricH]

I accidentally tip your robot over with your gravity-down lift up. Oops, your lift doesn't go down anymore.

330 always adds a "gravity" cable to lifts for the case where you're tipped over and you need to unblock the field. It also helps to retract the lift for other cases like hanging on a bar. It's not that hard to add to the drum, either, if the winch is sized right.

Scissors lifts can be really nasty to get right. If you do get them right, they work OK... but they do have their weaknesses. I've seen a single-joint arm clean game pieces off the top of a scissor lift with one sweep, back in 2005. Hit a scissors lift so that it sways and... well... not pretty.

330 does have one team rule: All deployables shall be retractable (exception--we decide it's not necessary to pull them back in). Reason: We don't want to have to be dragging something around all match because it deployed too early by accident--and it's really surprising when we pull up ramps and go defend a spoiler (2007) or do other similar things.

[Madison]

Originally Posted by EricH

I accidentally tip your robot over with your gravity-down lift up. Oops, your lift doesn't go down anymore.

What difference does it make that a gravity-lowered lift no longer works when the robot's been tipped over? Talk about not seeing the forest for the trees.

It's usually not difficult to power lifts in both directions and it's sage advice to do that whenever possible, but not doing so is not necessarily a catastrophic decision. There are all sorts of things that can be done to mitigate any problems that might arise from relying solely on gravity.

[EricH]

Originally Posted by Madison

What difference does it make that a gravity-lowered lift no longer works when the robot's been tipped over? Talk about not seeing the forest for the trees.

If a gravity-lowered lift is up, and the robot is tipped over, that lift is now blocking the field. Because it's gravity-powered, it can't go down (robot-relative) any farther. This can lead to penalties (the 84" cylinder, for example--a gravity-lowered lift that falls over while trying the top peg could easily be over 84" from the floor, causing a cylinder violation) or difficulty in getting the robot upright (longer moment arm, so more motion needed to go the same rotation).

If it could be pulled down relative to the robot (not the floor), then there is less chance for penalties, it's easier for a partner to get the robot upright, and depending on design, the robot itself may be able to assist the righting process.

There have also been some comments--some time back--that dirt could jam the guides, or some other such things, that would interfere with gravity.

Can you mitigate it? Sure. But Murphy says that the problem will happen when the mitigation isn't happening right.

[GaryVoshol]

Originally Posted by EricH

This can lead to penalties (the 84" cylinder, for example--a gravity-lowered lift that falls over while trying the top peg could easily be over 84" from the floor, causing a cylinder violation)

Nope - the cylinder falls over with your robot.

Originally Posted by <R07>

(note: these limits are defined in reference to the ROBOT, not the FIELD).

[thefro526]

Originally Posted by GaryVoshol

Nope - the cylinder falls over with your robot.

In previous years, this wasn't always the case.

[IKE]

As many have noted "Never" is a really strong word. Most posters got the intentions of this thread which were to talk about some ideas that should ahve worked, but had major issues in implementation.

[Al Skierkiewicz]

Although the games lately haven't brought out shopping cart wheels on the front of a robot, don't use them. They are especially bad when the game has an uneven playing surface. The robot goes where the wheels tell it to go so you can't climb a ramp with shopping cart wheels.
Don't ever believe someone who says "of course that's legal, we have been using them for years." That is why at least one team will be down hearted when I have them remove their Globe motors this year.

[ChocolateDragon]

Originally Posted by Dkt01

Our team will not likely use vacuums, or conveyor belts due to flawed execution and/or excess time we spent fixing parts. We will also never make a robot with less than an inch of ground clearance again (got stuck on the field repeatedly last year).
I agree that almost anything can be successful if done correctly, some ideas are just more difficult to implement than others.

What about the conveyor belt you tried didn't work? Was it the flawed execution, or the excess fixing time, or some other reason?

[thefro526]

Historically, Drive-trains have not been our strong suit.

In 2002, 2004, and 2005 we used 4 wheel drives using rubber wheels on all 4 corners in a long base configuration. None of those robots turned very well at all, and one of then had a tendency to leave the ground while attempting to turn...

In 2006 and 2007 we decided we were going to try out Omni Directional wheels coupled with IFI High Traction Wheels. For some reason, still unknown to me, we chose not to power the omni wheels and only powered the traction wheels... These didn't turn all that well either...

Since then, there have been two unspoken drive-train rules on our team: No long base 4WD and No Un-powered Wheels.

[Jared Russell]

Relying on gravity to do anything other than to give your robot weight. If it moves down (an arm, an elevator, a ball in a hopper), make it powered.

Ditto scissor lifts and "many-to-one" hoppers. I've seldom (but not never) seen implementations of each that were effective.

Mecanum wheels are on the verge of making the list, too (for the same reason).