That’s a good point! That’s why we’re just planning on dumping our preload low. Also based off of our bump side auto testing in 2023, the AprilTag should be able to help us get back on track more. How much? Well we will test it and make a post on it later 
9 Likes
Tilted Elevator Architecture
Similar to designs from the RustHOUNDS and Cranberry Alarm Ri3D teams, we discovered that a fixed CORAL ramp or hopper with a manipulator on an elevator could enable scoring on all four levels. However, our design has a CORAL manipulator mounted at a fixed angle to a slightly tilted (~10 degree) elevator, and is designed to allow CORAL to pass through the elevator for front-to-back or side-to-back scoring.
A tilted elevator lets the CORAL manipulator extend further from the back of the elevator than would be possible with a vertical elevator near frame perimeter, reducing the distance from a raised CORAL piece to the BRANCHES from around 12 inches to potentially zero, especially for higher BRANCHES.
Elevator in the L2, L3, and L4 positions
This design also has only one axis of motion for the CORAL path, with 2.5 total degrees of freedom including rollers. The simplicity of our 2024 robot was integral to our performance, having only 3 DoF including rollers and a climber, so we see only 2.5 primary DoF as a big plus over other designs like an elevator+wrist or elevator+telescoping tube. A few independent DoF will probably be added for an ALGAE floor manipulator and a climber, but they should be easier to manage than a wrist.
Post written by @Octagon89098
15 Likes
With this design, how are you thinking of removing algae from the reef? Our team is also really trying to not put a arm/wrist on the elevator, but at this point it seems like we have to with the need to extend out to remove algae.
2 Likes
Probably some sort of stationary roller on the elevator that would go under the algae and pry it out either by spinning the roller or backing the whole robot out. @Octagon89098 can speak more on this (I’m just a lowly programmer 
)
4 Likes
back to back posts woo
Jan. 6 Prototyping Update
We have mechanically finished the Alpha robot, and are almost done with code.
Hopefully it will be driving and scoring (maybe autos!) by Wednesday!
We also prototyped a horizontal roller claw for Coral acquisition/indexing:
It generally seemed to work well for intaking even without the front rollers powered, except when the Coral was inserted fully sideways at which point it got stuck. With the front rollers powered the Coral could sometimes get stuck between the two sets of wheels and needed to be removed but this could be solved with different geometry. For placing Coral this design did not seem accurate, and it would probably need to be combined with a separate scoring mechanism
4 Likes
i’ve replaced the photo with an embed, do the video links work?
(the videos should work now)
5 Likes
Yes they all work now!
1 Like
#throughelevatorgang
4 Likes
with the tilted elevator would you have to move the robot back (not be against the reef) when doing L4? thats one of the things that I saw when doing a tilted elevator
2 Likes
Two reasons we think we won’t have to:
- The L2 and L3 scoring will involve the coral having a little bit of air time. We’re allowed to launch coral if we’re in the reef zone, so we just have to send it and clear a gap of a few inches.
- It will help on L4 if the wheels are squishy enough.
3 Likes
We aren’t entirely certain yet. If possible, we will remove the algae without adding a powered wrist. A spring that deploys a small arm at match start with a wheel on the end may be enough. In fact, the motor that powers the coral manipulator could be used to spin the algae ejector.
4 Likes
Alphabot Lives!
The intake jams a lot, and we don’t have scoring positions set right (that reef we scored on in the video is not field accurate at all), but it drives and picks up from the feeder station position and can outtake. We will be using it for auto development and drive practice going forward.
In other news, we have finished wiring/assembly for our Beta chassis:
We also prototyped a slap-down algae grabber design and a coral roller claw end effector:
Both work alright for what they are supposed to do, and we will probably use something similar to the algae intake on our final bot since we only really care about knocking the algae off and scoring it in the processor. The roller claw could be useful for an outtake on a design with a separate intake like our slanted elevator sketch.
14 Likes
Alphabot Cycles
https://photos.app.goo.gl/eWkDa43E4wU11few9 Drive practice has been extremely rough, and we have learned some important things. We had to upgrade the intake in multiple ways, all with just gaffers tape (some quick sketchy spacers and a net to stop the coral from falling downward).
We also put on a front bumper (that actually came from our 2007 robot!) to better simulate lineup, and a secondary net to stop falling coral from landing on electronics after one unplugged our elevator motor (this gamepiece is scary when it’s falling!) . Overall the alphabot has many issues (spaghetti wiring, old motor firmware issues, can only reach l3, no centering, drops coral), but it has served its purpose as an early moving robot to try to figure out aspects of the game. We might throw a version of the horizontal V intake on it soon, but most likely it will serve as a drive practice and programming bot for the forseeable future.
10 Likes
More Alphabot Drive Practice
our drivers have been improving pretty rapidly on this bot already, but that isn’t all we’ve been doing. Programming and Design updates hopefully coming soon.
6 Likes
Programming Update: Auto Align
It’s pretty obvious, but the coral require a lot of precision to score especially since the visibility could be so low (just watch the videos posted by @Paul_R to see our struggles ), which is why to perform at a higher level, it’s almost necessary to have auto align or have a super insane driver. Since we only have an insane driver, our plan is to have a button that will have the robot to autonomously align with a predetermined reef (decided by our insane operator). The most apparent way to do this is to use the robot odometry to drive to the desired spot, except the issue is that the odometry always has some error and we want to be precisely accurate every single time.
The Plan
What do plan to do is to essentially split the line up into two parts: 
drive to the reef using robot odometry and then a ↔ 
precise alignment with the reef AprilTag using a wide angle lens camera (so that we can see the tag even when we’re very close to it). A lot like the reflective tape in the pre-AprilTag days, we plan to align by driving a certain direction until the AprilTag is in a predefined spot on the frame. This will let us know that the robot is aligned and ready to eject the gamepiece. We have been working on this with our 2024 robot the past couple of days and have made significant progress! Some photos and videos are attached below:
Camera set up for testing
Videos of the lining up
Hopefully in the coming days we can integrate this with the auto driving part. Look out for a post about that soon!
10 Likes