COTS: How far should it go?

[aldaeron]

I do not see what the big fuss is about. There is nothing revolutionary in the kits, with the exception of the WCP MCC (which I will come back to in a minute). Previously, people would look at old video of what was successful on CD or photos from previous regionals. Let's use the case of a roller intake for a spherical game piece. There are only so many ways to pickup a sphere off the ground using the actuators allowed in FRC. A "shaft full of wheels" or "tube with grippy material" on a rotary joint is a lightweight way to grab a compressible sphere without needing precise alignment (remember you are driving from 50 feet away). In the past, someone would ask for help or details of how teams did it the last time we had a compressible spherical game piece. Soon there would be robot photos, mostly on CD, and people could quickly figure out what parts were needed to build the design and order and assemble them. Now there is a convenient kit you can order with less clicks. The end result is the same, roller intake using the same principle as years past, with the exception that some teams who are less good at searching these forums are now aware of the kit and can buy it COTS. It is also important to note that just because you buy the kit, does not mean you have to use all the pieces or make it exactly as the photo shows it. You may choose change the arm dimensions, motor used, gearing or wheel size after testing.

Same thing goes for gearboxes. You are designing a set of spur gear reductions and maybe adding a chain sprocket reduction at the end. Spur gears need to have their centers positioned within a few thousands of an inch of each other. If you don't have a precise way to get the gears mounted correctly they will destroy each other. There is a still a ton of physics and design tradeoffs with COTS gearboxes (1 vs 2 speed, 2 vs 3 CIMs, Gearing ratios optimized for time to distance, pushing max current at traction limit, etc)

We could keep going into all the COTS parts, but the end result is the principles are the same and vendors are making the execution easier and easier. The seven simple machines have not changed, their execution for FRC has just become easier (assuming you can afford to buy them).

On the note of the MCC, it is tempting to say that this will "break FRC" because you can purchase a robot. Let's be clear you are only purchasing the mechanical components, not the software, ability to drive or game strategy. You could give identical robots (with identical software) to teams and it would still be a ton of fun to watch. Remember that just because the MCC or KOP kit does something one way, does not require you to do it the same way. Maybe you change wheels or gearing ratio or the size of your roller intake wheels. Keep in mind that with publicly released designs (MCC, Ri3d), that everyone else sees them and is looking to design something at least as good as those designs, if not better. You can also design defensive strategies that counter these publicly available designs.

Personally I would like to see 20 MCC designs ready to go at the competition with working software and instructions on how to use them so that teams who tried their best and for whatever reason were not able to make a bot that can score points would get to participate in the C part of FRC (competition).

I am going to link to an old reply of mine on essentially the same topic since many of the points are valid. It also shows how many times this discussion has come up (though it is good to revisit it once a year).


Final note - as a team that used to have access to CNC machines, it was nice to be able to get into the details of fabrication with students and make custom gearboxes (even if COTS was cheaper and lighter). Now that we do not have access to these tools, we focus our attention from fabrication theory (i.e. climb vs conventional milling, surface speeds that give good finish, how to cut various materials and how to efficiently program different milling operations with the fewest setup) to constrained design theory (we want to make motions like X,Y,Z that fit in area A and we can chose from Q set of parts that we can afford from vendors 1,2,3)

-matto-