I’m happy to answer whatever questions you (or anyone else) might have about this concept. The original thread is a good knowledge base, but not all encompassing. It’s cool to see people pursuing this design. It’s a bit off the beaten path.
Could you guys get the carbon arm cad sent over! We have one working but it would help a lot!
UPDATE #3!!
I apologize for late update as I was fighting my demons studying for the october SAT and my team has as many students as I have fingers on my hands 
There was a lot that was accomplished since the last update so lets get right to it.
ARM
We had already proven that conceptually it would work out and it was just a matter of giving the arm some nice plates and a solid mount to connect it to the chassis.
On the software side I’m waiting for this to be mounted onto the chassis to begin tuning because I dont trust the results I’d get from tuning this on a makeshift clamp setup.
CLAW
This centerpiece of our robot has also seen some extreme refinement in the last week or so. The first major change that we made is that we rubber dipped the rollers because we noticed that the biggest problem we had is that there would be too much initial slippage when trying to pick up a piece.
We were wonderfully pleased with the rubber dipping results as not only do we have more grip it also meant we can pick up from the floor while applying less pressure. Heres a video of the ground pick up in action and a test using it to pick up a specimen and hang it.
Additionally, now that the the wrist attachments to pitch and twist the claw have been made I was able to put my money where my mouth is using opencv. I present to you a claw that auto aligns based off the samples orientation. The orientation of the sample itself can be calculated relatively easily using info from the bounding rectangle and the actual logic for mapping servo position to detected orientation was handled by a linearinterpolation class I had stowed away from rapid reacts. Obviously this will need to be optimized but I’m glad I was able to get the rough concept working well enough.
FULL ROBOT
After far too many allnighters our cadding duo have arrived at a somewhat finalized design for our mk1 robot.
You can see how everythign sort of comes together here and we after long hours of machining over the weekend we got all the pieces cut and ready to assemble.
Wrapping things up I’d like to provide the link to a folder that contains all the files for the claw. I apologize for the delay we’ve had in posting it but please keep in mind we only had 1 student cadding this and between getting the cad files out of him and filling out my college apps I couldnt be as punctual as I desired.
Next on the to do list is to make a similar folder for the arm related stuff and to update the github with the alignment code which will happen after my calc test hopefully. Cya
2 Likes