Field-relative positioning through non-linear state estimation
23 Likes
Do you ever see going back to Colsons in the middle? Not counting games like 2016.
Why the switch back away from mag encoders on drive?
Was there any damage to the thin wall tube on the arm?
Why dog shifter over ball shifter for the drive?
2 Likes
I’ve heard that top teams refrain from using colson because they slip in auton or something and so auton is less accurate.
1 Like
254 and a few other teams have noticed differing odometry depending on the direction of robot travel relative to the carpet grain with colson wheels. It was significant enough to warrant swapping back to treaded wheels.
4 Likes
1 Like
I know that part, and I remember Jared’s post from when he first wrote it. I was just curious if it was something that would become their new standard. Other than the added cost the only instance I could think of would be a field that was majority HDPE.
How does the bearing setup in the old stinger climber work it almost looks like their are ball bearings arranged around the outer tube.
To anyone: Blue nitrile vs colson in detail
Brake system- how did it work?
So the way 254 does it is that they use an over-centering linkage which hits a stop slightly before the piston hits the end of its travel. You can see more details in this post here.
Was the acceleration gains from two speeds with neos worth it looking back on the season? Why not chain in tube for the drivetrain?
1 Like
Has 254 ever done chain in tube?
1 Like
yes i believe in 2015.
I’m curious about the suction climb and what happens after the match ends. I know there’s a ratchet so it doesn’t fall down, but that still requires a vacuum seal on the platform to work. When the robot is disabled the vacuum pump is no longer running. How do you maintain that seal for extended periods of time even after the match has ended? Is there some sort of tank to extend the volume of the seal? Or does the pad work on it’s own? Is there some sort of safety mechanism on the chance it does end up falling down? How long could it theoretically stay suspended on the platform?
3 Likes
They use a check valve. It allows air to be pulled out but not let back in.
It’s still a valid question. The seal isn’t perfect and air is slowly leaking around the seal into the volume at vacuum. It’s only a matter of time until suction is lost and the robot falls. Maybe it’s 1 minute, maybe 1 hour.
3 Likes
The seal will actually hold for over an hour for a weighted robot with no drop in pressure from our tests in our lab. After an hour with no pressure drop we released the pressure but we deformed our foam padding for a good couple hours before it returned to its normal shape.
If you bolt through the suction plate you need to make sure you use RTV silicone or something similar to seal the holes.
During our test we started kicking our robot on the side to see if we could pop it off the vacuum and it ended up sliding on the surface with a good whack but still held. This was after the robot was disabled and the vacuum stopped however the check valve held the vacuum.
In the end of the day it comes down to execution in manufacturing as to how well it seals
10 Likes
That’s awesome! I never imagined it would have been that long.
What was the condition of your test surface? I bet the seal to a new piece of textured hdpe would be better than to one beat up over the course of a competition season.
We tested on a new piece of HDPE. The seal of the vacuum with the closed cell phone really conforms to the surface. Even with large gouges on the surface of the platforms during later regionals and at champs we never had an issue. We did have a couple drops at our LA regional but that was from someone forgetting to put RTV silicone on the bolts attaching the vacuum plate to the arm.
1 Like
It was quite tricky because we had not planned to have a suction climber in the beginning, we just thought it wouldn’t be that easy to hold suction on the textured HDPE and didn’t really do much testing to pursue it. When we saw the success of 1619 among others we started to investigate and realized it wouldn’t be impossible. It was hard to package such a long leg and folding linkage system into a narrow width volume (when stowed it needs to fit within the frame perimeter).
Luckily, the climber carriage system was pretty simple, basically inverting the bearing setup compared to the Stinger climber. The stinger climber had a short fixed outer tube with Teflon sheets on the inside that the inner tube slid through. The suction climber has a tall fixed inner tube and then a carriage made of plates and ball bearings on screws that slide around the inner tube. Both fixed structures mounted to the turret gearbox plate, with the suction climber upright also mounting to the crossbar at the top of the elevator for added strength.
2 Likes