When to make practice bots

[lnader]

Here's our schedule for the last two seasons:

Weeks 1 and 2:
Hardcore design and CAD work, resulting in a full, functioning robot model and part prints. Mock up kit chassis for programmers to start tinkering.

Weeks 3 and 4:
Fabrication, assembly and testing of practice robot. Noting design changes as we go.

Weeks 5 and 6:
Fabrication and assembly of competition robot.

The definite (and as I feel most important) advantage to building a practice bot is drive practice. Confident drivers make less errors which results in better performance on the field.

You could also say that it gives programmers more time, but you could have a 12-week build season and the programmers would still want more time. (sorry Tubby)

[mathking]

This is the first year in which we have built a true practice bot, in other words one which is mostly the same in form and function to the competition robot. In previous years the practice robot was just a drive chassis or was significantly different from the competition robot.

This year we built the practice robot (Scotty) first, because as several posters have mentioned, it allowed us to focus on getting it running quickly and testing its components. The final robot (Mike) is somewhat different. It is an inch longer (27" vs. 26") and 4" shorter than the practice robot. The longer came because it added stability (we could move the battery further behind the rear wheels) and made the "comb" to keep the round urethane belting from walking easier to implement. The shorter came because testing on the conveyor/garage door sensor system showed it was more precise than we planned for, so we didn't need as much space as our original design indicated. The down side to these changes is that the practice bot, while very similar, is not an exact copy. The plus side is that we were able to make the kind of improvements on the design you can only make by testing it.

This process also prodded us to make the robot design more modular. The shooter/turret assembly, the conveyor and the ramp manipulating arm are all attached basically by four bolts. Take out the bolts and disconnect the quick disconnect wires and the parts come out. In the end we have built two of almost everything.

We assembled the final robot in the last week. Next year I am going to push to get it done a few days earlier than the final weekend, but the trade off has been worth it. Scotty is a fully functional robot with the same basic design as Mike, so we have been able to practice. Over spring break we will get in quite a few more hours of driving and calibrating. (Just have to remember to move the basket height up on our backboards to account for the 4" height difference.) This is particularly good this year, because both drivers are rookies to FIRST, not just rookies to driving. df

[Dale]

We use our practice robot as a way to get all the kinks worked out of our design. We don't have the skillset and time to CAD the entire robot in the short build season so we just CAD the rough overall plan and CAD particular tricky parts. The rest we work out on the fly with the practice robot not worrying much about weight. We use 80/20 for the practice robot because it goes together so much quicker and we can change our minds. That's usually turned over to software by week 4. We then switch to 80/20's Quickframe and other lighter materials for the competition robot which usually get's done half way through week 6.

When things like the drive train are set in stone (usually around week 2-3) we'll start doing that portion for the competition robot in parallel with higher level things happening on the prototype.

[Jared Russell]

We have found that the greater the delay (in either direction) between the construction of the practice bot and the competition bot, the less the two bots look like identical twins and the more they look like fraternal twins.

[jvriezen]

We went from never having built a practice bot to building two this year (sort of)

Our mecanum wheel modules in 2011 and 2012 were similar enough that we could take last years modules out of our production bot and put them in a plywood frame (last year was narrow bot, this year was wide). The plywood frame effectively matched our production frame (which by the way is made of plywood and won the Industrial Design award at Lake Superior). We added 75lbs of weights and our driver practiced with it.

We also built a non-mobile base to mount our withheld shooter assembly so we could practice targeting and shooting.

Unfortunately our 4 slot cRIO went south as did its replacement from NI, so we had to share cRIOs between the two practice bots, so we could only use one at time.

[Koko Ed]

We design the prototype before the final bot so we can find out what works and what doesn't and take everything we have learned from the process and design an even better robot for the final competition bot.