This has been in the back of the oven for a while. Hopefully it isn’t too burnt.

A long time ago someone made the “encyclopedia robotica”. It was part of my introduction to the world of FRC, and I liked the idea.

It isn’t super comprehensive, though, is dated, and I prefer general-use materials that apply beyond our little robots. So… I do what I always do… roll my own. I also put it in LaTeX on GitHub so that it can be kept up to date by readers like you as the kitbot becomes a differential swerve and the BAG suffers the same fate as The Bag. Any feedback or assistance is appreciated.


(Technologies, Principles, Design, and Analysis of Complex Electro-Mechanical Systems)


This is a great read. Highly recommended especially for those that are just getting into FRC

If you want some pictures of FRC spaceframe chassis my former team used to do those


I think I’m gonna try tossing this at a few of the mechanical students this year and next as a sort of experiment. On its own, not only do I think its super useful, but I would think there’s a huge value in seeing just the list of vocab words, terse but reasonable definitions to point in the right directions, and pictures of the things. Breadth, not depth, is the name of the game when ramping up a new FRC student.

Many congraduations and +1’s and virtual virus-free high-fives to Thad Hughes!


What a wonderful resource! In my FRC experience there has always been so much vocabulary to learn, and this explains it all!

Small note: I think the crab and swerve image labels are switched in section 8.5.5

Amazing resource! This is definitely going in my archives

I see that in 10.7 the error is added, which would be correct if the error was a bias. How would you add the error if it were uncorrelated?

1 Like

This gets into statistics…

If the error is fully engulfing (e.g. 100% of parts adhere to a +/- 0.1mm window, not necessarily following a normal distribution) and you’re looking for worst-case scenarios, it stacks as I wrote.

If the error is a standard deviation and you’re looking for an average error, it scales roughly by the square root of the sum of all error components, squared. (Full story: Propagation of uncertainty - Wikipedia)

I am not a six sigma black belt or whatever you ninjas are doing today… but…

In some engineering contexts +/- 0.1mm would mean “all parts will lie within +/- 0.1mm”; and you want to be sure that all assemblies from said parts work, so look at the extrema and make sure that they do. E.g. if you want to be 100% confident, consider the worst cases.

If you can accept a certain failure rate or yield, (and you’re checking for it), you could use the error propogation formulas and arrive at a more liberal estimation.