Robot Tipping issues

[GeeTwo]

We haven't done enough driving yet to say for sure, but once we decided on an outside lift, we made our wheelbase as long as we easily could (KOP long chassis, 4" wheels). We expect to have 30-40# of ballast (plus the battery) at the bottom and back of that length to minimize this problem. If things are still problematic, we may cantilever some of that weight outside the "frame perimeter".

[cglrcng]

Here's how you fix that balance problem...LOL....That tells me lots. Balance out the weight, just make the bot do a lot more!

https://www.youtube.com/watch?v=8xiXzyIDXYw

Something tells me they will be in demand. Though once the big walls go up, they will be driving very carefully! I saw a stack of 6 capped in under 70 actual seconds w/ very lil' wasted movement. 4 at a time from the landfill, I'd be doing a layer of 4 in 2X2, go get 4 more, and 4 more, and layer the bottom all the way across 4 high~5 high first. Allow the others to cap and stack 1~2+Bin+Litter. They hit upon a really decent design w/ little wasted moves and really sturdy w/ a low CG it appears even though at max height.

[MrJohnston]

We seem to always make "tall" robots, so "center of gravity" is an issue we address yearly.

* If you are going to have weight up high, consider a wide base.
* Put every bit of weight you can as low as you can... Attaching weight below your axles works wonders.
* Talk to your programmers: Most tipping happens with quick starts and stops. If your programmers can control your acceleration, you won't be nearly as likely to drive your wheels out from under yourself.
* Know your robot's limits.
* Weight symmetry: Make sure that your center of gravity is truly over the center of the robot.
* Talk with your alliance partners. The last thing you want is for them to get careless.

[Abhishek R]

Originally Posted by MrJohnston

* If you are going to have weight up high, consider a wide base.

I would think you would want a long base, so that if you accelerate quickly forwards then back you won't tip yourself easily?

Though I guess if you get swiped from the side with a long base, you may risk falling over sideways, though I haven't seen that a whole lot.

[MrJohnston]

Originally Posted by Abhishek R

I would think you would want a long base, so that if you accelerate quickly forwards then back you won't tip yourself easily?

Though I guess if you get swiped from the side with a long base, you may risk falling over sideways, though I haven't seen that a whole lot.

Big footprint.... Not skinny in any direction.

You have to also consider turning... If you are long and skinny, you might be okay with a quick acceleration. But if you need to turn, you could topple.

[JohnSchneider]

Originally Posted by Abhishek R

I would think you would want a long base, so that if you accelerate quickly forwards then back you won't tip yourself easily?

Though I guess if you get swiped from the side with a long base, you may risk falling over sideways, though I haven't seen that a whole lot.

Just use your robot to right itself

[ratdude747]

One of the oddest tricks I remember was in 2008. My team at the time (1747 HBR) had a problem with overshooting when hurdling the overpass which would snare the grabber on our angled forklift and tip us over to the point of being caught on the overpass with the wheels off the ground. Our fix was to (IIRC) zip tie a piece of 3/8" round stock to the lower rear frame rail, which worked as a makeshift wheelie bar and prevented the tipping issue from disabling us.

[MrForbes]

We did the sideways tipping test last night, our robot has to go to about 45 degrees leaned over before it will tip over.

Still need to do the front and rear tipping tests, unloaded and loaded. Although we kind of have a safety, since the tote stack falls over way too easily.

[Batterink]

We did two things to prevent tipping. We designed as much wait opposite to where we hold the totes/cans as possible. That wasn't enough, so we added weights (dumbbell plates) to the back. We also noticed that if the driver accelerated, or "decelerated" to quickly we could either tip, or lose our stack. So we implemented "ramping" code that limits the acceleration of the bot.

[Jon Stratis]

We're about as tippy this year as we've ever been... it takes about 65 degrees before we'll tip over this year. We've had that number as high as 80 degrees in the past!

[Trey178]

We've designed the top deck of robot with the ability to slide back via a motorized lead screw. When we lift up 3+ totes above the top of our frame, our operator will slide the deck back, moving the C.O.G. with it. It will move forward when we're ready to score.