This robot truly is amazing. Is it ok if I take some inspiration from it for my team’s robot.
“Steal from the best, and invent the rest” -Michael Corsetto, team 1678.
Go for it, and good luck with your robot.
Excellent reveal. I’m predicting this design with slightly more stable lifts for stacking on stacks and bigger trays in the finals of worlds.
Thanks! I don’t think this bot design type has drawbacks aside from weight and stability. the lift adds a significant amount of weight to the robot, and takes motors away from the drivetrain.
Exactly, and they’ll get lighter and more stable as the season progresses, and maybe someone figures out how to make a viable 1 motor intake to support the 4 motor drive.
Nice robot and reveal! I particularly enjoyed seeing your tray deploy - having cubes push the hinged tray up is an unusual design but I think it has some strong advantages over the more typical design (i.e, tray flips out when the lift goes up). Most notably, the tray flip-out won’t be as abrupt. I’ve seen some trays that expand so fast that they extend out of the field and almost hit their drivers, no way that’s legal under S1 
I was a bit surprised you opted to go with a 3-motor drivetrain and 2-motor intake instead of a 4-motor drivetrain and 1-motor intake. Was a 1-motor intake too complex to design on the lift? Did you need the power of 2 motors to hold all the cubes in the tray? Having a 3-motor drive is a significant detriment when it comes to defensive driving, which we might see a lot of in Tower Takeover.
On a related note, the strafe wheel (H-drive) is an interesting addition, especially considering you’re using a roller-style intake. I don’t think we’ve seen any traybots with strafe wheels yet. What were your reasons for including it? I’d imagine you could use the V5 motor “hold” function on all 3 drive to keep your bot from getting pushed at all; this might be a good idea when scoring in the unprotected zone.
The slide-out wheels at the back are a cool feature. Have you had issues with this design tipping?
You mentioned you couldn’t stack atop existing stacks. Because of this, do you see any advantage to lifting up the entire tray when scoring a tower, instead of just lifting the intake? It seems you could have the tray fixed to the ground and put the intake on a DR4B, 2-bar, or other sort of lift, so that when you score a tower, the bot lifts up only the cube it will drop in the tower. Were there other reasons you had to go with a tray-on-lift?
Thanks for sharing the reveal. I know I bombarded you with questions, if you don’t want to answer all of them you don’t have to
I will answer when I get home
I’m not on the team, but this is what I think:
1m intakes on traybots have been tested, and usually don’t tend to stack enough cubes. 2m 200 rpm has generally been found to be the best middle ground between speed and strength. Especially if they can’t stack on stacks, this robot should be able to make stacks of significant height to be viable.
H drive is an interesting design choice, especially with the versatility of the goofy-style intake. I assume that it has been added to assist with tower scoring: I believe we’ve seen h-drive on other tower bots as well. Time will tell whether it’s worth losing the power.
As for the tray-on-lift (schlong), I believe the main reasoning behind their design is shown in their reveal. They show their bot removing opponent’s cubes from towers then scoring their own: this wouldn’t be possible with the standard design of a goofy dr4b.
Sorry for the late response, I am currently rebuilding Usagi to have a wider lift and thinner base to fit into the wide scoring zone easier. On a side note, would completing the robot within 3 days count as a rI3d even if those days are not consecutive?
Thanks for the compliment on the tray. I actually disagree with you, trays should be deployed without cube interaction. I know from experience that trays with more sections can’t rely on the cubes, else they run the risk of blocking the intake and ruining an autonomous period. I’d much rather the intake hold the tray in and the tray hold the intakes up. It’s not much of an issue, as 4442X (who have a tray deploy which is held in by rollers) simply warned me to not stand behind it lest I get nut tapped.
Yes, the 1m intake was too complex. Each side is independent of each other and folds down. I wasn’t able to find a feasible way to chain both together reliably. I did not need both motor’s power, one probably would have been fine. 3m drive is fine, at least in Hawaii where there isn’t much defense. I used 3.25" wheels, and at least for me, a mediocre driver, it feels similar to 4" wheels. The top speed is less, but the acceleration to get there is present. It’s not too much of an issue, though, as there is a lot of delicate work involved in stacking. It plows through prestacked cubes with ease, and allows me to gently move stacks to fit in the wide goal.
I wanted strafing to maneuver around the field easier as well as dealing with towers. I think that the strafing wheel gives a lot of maneuverability to get around easier. I’m not so sure about the drive hold idea. I’ve experienced the motors starting to heat up and run out of juice after a while driving.
I have had issues with the design tipping. The tray is heavy and the bot is liable to tip backwards when the lift rises up. Making this issue worse is the drive base shroud on the back of the chassis, which further moves back the center of gravity.
@Adam_14606A is correct.
The main thing with lifting up the tray is the speed of swapping. Anyone can plop a cube into a tower, even clawbots. A few less can quickly remove those cubes from towers, especially if they are at an odd angle or are nested. Few if any clawbots can “cache” their desired cubes so that they can quickly swap without triggering a response from a defending bot. I am able to suck up that cube, spit it out, then spit out the desired cubes in the time it would take for say a tube bot to grab the cube. It saves the time of lifting down and up, as well as hunting for a cube and realigning if you have a non linear lift like I do.
The main reason that I went for the lift tray is that swapping speed. I don’t mind the downsides of a heavier bot and weaker drive, else I’d take a motor off lift. Caching cubes in your intake and being able to score towers consecutively is where this robot shines. 4442X has a lift tray and I don’t think they would have been able to hit 112 without it. They were able to quickly stack their high stacks and hit multiple towers in a line to the next. They hit all towers, which I am not sure a two bar traybot would be able to do.
Wait, a DR4B is a linear lift. Did I miss something?
The bottom 4bar in my dr4b is about 14" axle to axle, while my top bar is 17" axle to axle. This, as well as the gear rotation offset mean that my dr4b pulls back a bit at the top of the lift curve. This makes it inconvenient to score cubes in towers.
That’s a really interesting strategy I hadn’t thought of before. I had initially predicted that descoring and scoring of towers would all be done separately. As you said, directly swapping brings a huge improvement in speed and reduced susceptibility to defensive tactics. I’m excited to see how this strategy plays out during the season.
Thanks for taking the time to share more details about your robot! Good luck in competition 
I’m happy to share. Good luck to you, too
Lift linearity isn’t truelly much of a concern. The affects are minimal… although for stacking on stacks even the small things matter. Linearity might become increasingly important as this design evolves for stacks on stacks.
It seems that tray bots always work in Videos and they make everyone want to build them, but when they actually go to competition the stacks just keep on tipping over.
its all about practice.
The driver of the traybot needs to get a lot of practice to put the stack nicely into the zone, while under pressure.
But traybots do work…
I feel that this is because of a handful of reasons.
*Tray tilting is too sudden / jerky, causing stack to fall.
*Release method jerks or moves the stack, causing stack to fall.
*Tray tends to lean back towards the top, meaning that the top portion of the cubes lean into the tray and sort of push the bottom cubes out.
The solution is simple for each of these.
Use P control. Proportional control basically ramps down motor power as the motor reaches a rotation target. Ex. threshold+100(1-(motor.rotation/tilterTarget));
threshold is necessary to give the tilter a little bit of extra leeway in terms of tuning. Unless you are being super precise it’s hard to return the tilter to the same location each match.
*Use shorter flaps. These have a little more margin of error compared to longer flaps in my experience. Additionally if you tune your intake you can create a function to outtake as you reverse.
*Use a C-Channel on the underside of your tray extension to align both tray sections.
I forgot to mention: Make sure to untilt. This allows you to check if your stack will stand and avoid your tray hitting your stack as you reverse.
It’s not too hard in my opinion. With P control you can just roll up to the scoring zone and hold a button till the tilter stops, then untilt and back up.
well… i know my teams did used some sort of control for the tilting during singvex.
so you are right over here.
V5 has built in PID! -and it’s actually pretty good