Hi everyone, my name is Kanta, from Japanese team 8853.
This is the second time we’ve ever participated in FRC, so we have so many things we don’t know, and only three team members are making robot. So I want to ask you some questions.
1 Designing robot
I don’t really know the procedure of building robot. We’ve discussed the structure of the robot in our team, but we’ve not decided it yet.
Can you tell me what to start with?
2 Details of structure
I suppose most teams start their making robot with using CAD, and we’ve practiced using CAD during off seasons. But when it comes to designing robot on CAD, it was more difficult than I expected. I’d appreciate it if you could tell me the process or tips when you use CAD to design robots.
3 Division of Roles
As I said before, we have only three team members who are building robot. Since we are using Onshape as our CAD software, we can share CAD files and design together. Should we design robot together or should team members be assigned to each part of the robot?
As others have stated Everybot is a great starting point (just pay attention to keeping your mass low). Everybot has superb documentation, thus the recommendations. From a multi-season learning perspective it will be useful to compare and contrast against a lot of very similar machines to learn some of the nuance of what works and why. Just to reiterate, KOP drivetrain is a great piece of hardware. Some of the other open alliance threads are great for spurring discussions and iterations as well.
You may be surprised on how many teams don’t, or at least how many teams depart significantly from what is in CAD vs the bot. You can do a lot with a careful hand drawing, square, drill, pop rivets, center punch, step bit, band saw and a steady hand. No CAD involved. You just need to be consistent and methodical (especially with shafts and bearings). Traditional clock makers achieve amazing tolerances with hand tools.
I would divide things up a little so that one team member knows one part of the design extremely well, another team member know another subsystem, etc. It’s key everyone is still aware of the greater robot context (i.e. have meetings, talk through things, no one should be working in a vacuum). I would split it up along the lines of game piece manipulator(s), manipulator positioning (elevator, everybot arm/superstructure, etc), and drive train. But don’t get so lost in CAD that build time and crucially programming/ driver practice time suffer.
One last thing, base you design around what you can actually get parts for. I can put 20 falcons on my robot in CAD but I can’t order them. This year is far worse for parts than most. Substitutions for metric and part availability are going to be common.