Team 2102 Presents Our Offseason Robot


Team Paradox is proud to present our offseason project, ClawkBot!

Unlike years past, where we’ve built a mechanism or robot in anticipation of using the design in the next build season, we chose to build a robot to play this year’s game. We also decided to build a simple(er) and hyperspecialized robot, rather than building a robot that can play every part of the game (as we did for our competition robot). This robot is designed to be an efficient extremely switch/exchange bot due to its reversible intake design.

Here’s some info about the bot:


  • 20"x20" frame (without bumpers)
  • 72.4 lbs.


  • 2x2 wheel configuration (2x6" blue nitrile wheels, 2x6" omni wheels)
  • ~16 ft/sec theoretical top speed
  • Long layout on custom single speed gearboxes allows for completely gear-driven wheels (no belt or chain)
  • 3 CIMs per gearbox (6 total)


  • Arm can swivel a full 180 degrees to intake and outtake on both sides of the robot
  • Madtown/Citrus Circuits/Cheesy Poofse-style intake (modified design to function without pneumatics)
  • Spring-loaded intake with adjustable holding force
  • 17:75 reduction on custom intake gearboxes (1x775pro per gearbox)
  • 30x2" silicone rubber flex wheels
  • 7:375 reduction on custom arm gearbox (2x775pro)

The CAD for the robot and its mechanisms will be made available soon(ish), as well as the CAD for our 2018 competition bot, Lambda.

This picture isn’t very helpful for understanding the individual mechanisms, so let me know if you have any questions and I’m happy to answer them/post more detailed pictures.


Nicely done! Over what period of time was it designed and built? How many hours would you say you have into it? Do you have a photo without bumpers?


Awesome!! I love the way you’re doing the arm rotation! Could we see some closer shots/cad/documentation of that setup? Great robot, very well thought out, I’m so happy to see it all come together!!


Nice Bot, I’m curious do you have/plan to test the bot at an off-season event, or is it just for practice/teaching. If so what event are you going to try this bot at?


Here’s a shot of the robot without bumpers:

We started designing the robot at the beginning of the summer and met ~2 days a week for about 3-4 hours to work on the design—some work was done outside of meetings too, obviously. Everything was machined and assembled starting in very early September, and we worked on that for around 4-5 days on average per week for 3-4 hours per meeting. We completely finished it last Wednesday, which was just in time for Battle at the Border in San Diego, which we competed at.


The main goal was to teach new designers, as we lost a couple of our designers to college last year. We entered it in Battle at the Border, where we were unfortunately knocked out in the quarterfinals.


This is a shot of it in the mid-assembly stage. The gearbox drives an in-bar belt, which turns a gear that rolls along the plate gear.


The plate gear is for keeping the arm in a “locked” position when not being held by the motors?


This is really awesome, I’m excited for the CAD release. Thanks for sharing.


No… The plate gear is what the smaller gears drive up and down to move the arm from side to side. There is a belt going from the motors at the base of the arm up to that gear.


So the polycarb one and the white things on the wrist?
Really cool just havent seen it before.
Whats the benefit of doing this instead of just gearing at the pivot point?


Team 4613 did it as well

I think the benefit of doing this is that you get that much more torque without having to have a large gearbox, and the load of the arm is being held higher on it rather than just the very bottom.


With this arrangement, there is no torque being transferred through the axle/pivot so the axle/pivot only needs to support the weight of the arm and the intake mounted on the end. The torque on the shaft holding the gear that rolls on the plate gear is reduced from the torque that would be needed at the arm pivot by the ratio of the radius of the plate gear divided by the radius of the gear that rolls on it. From the photo in the 7th post, it appears that the torque is reduced by a factor of about 7x.

Reducing the torque prevents the damage shown in the photos below. The large gear was used to drive a small sprocket (12 ~ 15 tooth?) that then drove a large (54~60 tooth?) sprocket bolted to a our Stronghold arm. The arm had a similar range of motion to the one shown in this thread. The small gear was on a VP driving the larger gear.

What broke our gears was not the static torque needed to move our arm but the shock loads as the arm bounced up and down while the robot drove across the field, tarting and stopping and going over the Defenses.

Based on what the ME’s I work with told me, the shock loads can easily be 5~10 times the static load, depending on the nature of the shock related to a particular application (how many G’s). This would easily exceed any normal design margin that only takes into account the static loads. A Stronghold robot would not be considered a low shock (G) application.



This might help clear up confusion about what’s going on at the top of the arm assembly:


My image decided to delete itself, so I’ll put it up again…

We liked the design with a gear track because it meant we needed a great deal less torque to turn the arm quickly and still be able to hold positions.


Great minds think alike (although I think yours is a much better designed robot). Clyde will compete at Ruckus and the STEMley Cup (the latter driven by pre-rookie team 7475 wired)

5406 clyde small.jpg


You guys managed to package a cool climber into yours! That’s impressive, we didn’t think we could pull it off.

So, the reason that we chose that driving mechanism for the arm is that we saw 4613 do it and thought it was super cool. The fact that it took a bunch of torque off the pivot axle is a great bonus though. Originally the big gear was going to be 1/4" aluminum as well, but we got sticker shock looking at the price of it. Also, we didn’t want the two Al gears to gall together.

The original plan was to build two of them (red and blue), but given that the second one didn’t get accepted to Battle at the Border and how late we ended up finishing the red one, it’s a good thing we didn’t. We’ll probably put the second drivetrain together anyways. Good rookie practice.

Here’s a few extra pictures showing the guts a little better.


Hah, don’t get too impressed. We didn’t quite get there and made the difficult choice to pull the climber off for Ruckus. We might push to get it on again for STEMley.

We did look at having a 2 cube, 40 ball auto though :slight_smile:



Hey, it’s 120 balls or nothing this late in the season!

I actually kind of want to put a little turret on our extra drivetrain. Just gotta convince some students to design one.


One of the interesting discussions was the choice between one arm and two sector gears or two gears and one central arm. I think the students made a good choice in the final product that you see. The “little white circles” are flanges for the pinion that resist side hits and keep the gears from flexing sideways.