1) CAD is the design of the robot. Without CAD, there is no machining to be done, therefore no robot. On that note, without machining there is no robot and I've been wasting my time on all of these designs. Every job is crucial and the students I work with understand this. There are always going to be kids who are genuinely not interested in the slightest in doing CAD of any shape or form because it can be tedious, and staring at a computer for 14+ hours a day during the build season can be mentally exhausting. But there's nothing like designing a part and running over to the machinists and helping them cut the part. The job is a reward in itself.
I think taking a complex part and showing it to someone and explaining the difficulties that arise when designing or machining a part and how these problems are solved can intrigue that person enough to want to learn. Just don't make it sound too scary.
2) Most of our CAD members are sophomores and up. Freshman can be part of the CAD team if they want, there isn't anything stopping them. First year students don't always know what they want to do, but they learn the terminology and the concept behind the building of a robot. It's hard to teach someone how to make a fillet without them knowing what a fillet is. If a student wants to try and learn CAD we hand them a plate, a steel ruler, and a pair of calipers and tell them to draw it. Draw it again and see if you can find any tools to help speed up the process, or make it easier. It's pretty effective because by the time they're done they're familiar with the interface of whatever program you use, and that is about 60% of the whole process.
If they have questions, try to walk them through it verbally. Let them try and figure it out. When a mentor is working with me on the design of part they just sit there and advise me to do something or not to do something. Stuff like "that hole needs be moved" or "the OD of that is too big". Students work together with or without mentors to make sure each part is going to be strong enough, that it meets all of the requirements you need, and most importantly that it is actually machinable. Chances are, you'll revise that part 6 or 7 times in CAD, and cut it (hopefully) 2 or 3 times on the machines. It depends on how crucial and/or how complicated the part is.
CAD requires a concept for the robot. If you were to tell me to start designing the robot the day after kick off, I'd laugh at you
The first two weeks of the build are slightly boring for the CAD team because different concepts are being tested and we're not actually designing anything yet. At most we're hand-drawing basic designs/ideas. I make the best of that time and work on whatever i can get my hands on because i know that i'll be staring at a computer screen for 12-18 hours a day in a few short weeks.
Also, pressing Ctrl+S whenever you zoom in or out. There's been several times where I've wanted to take my computer to our table saw because I've lost parts due to a blue screen or a frozen CAD program.
Most of this is pretty common knowledge and common sense, but not always. You have to form good CAD habits! Everything depends on the situation of your team, and every team is in a different situation. It depends on your machining methods as well. You'll adapt as a team to any new problems.
3) We have a pretty small CAD team so things always seem to flow smoothly. It's just a matter of communication and knowing what everyone is working on. I don't have much to contribute to this question. However, I strongly feel that only one person should be handling the final assembly. I've had to redo entire assemblies because they become inconsistent and confusing when too many people work on them.
Another thing i'd add is to make sure all of your sketches are fully defined. I'm not sure what program you're using but Solidworks does everything with relations. If there's a blue line somewhere, that dimension is subject to change. I've spent hours on end trying to figure out why my part has suddenly become over defined or broken and then feeling really stupid when i realized what the problem was.
