An idea for the 2000 games..(is six weeks enough?)

[archiver]

Posted by Dodd Stacy.

Engineer on team #95, Lebanon Robotics Team, from Lebanon High School and CRREL/CREARE.

Posted on 7/29/99 2:50 PM MST


In Reply to: An idea for the 2000 games..(is six weeks enough?) posted by colleen on 7/28/99 1:26 PM MST:



Good ideas and discussions all on this thread. Here's my slant:

1) FIRST makes available on the web the list of kit materials, component performance specs, additional hardware, etc., and the manual sections relating to the RC system and wiring up the electronics from LAST YEAR'S GAME at the time of registration for this year's game. This gives the rookie teams an opportunity to familiarize with the 'typical' components and to come up to speed on the mystifying RC stuff, STAMP programming, etc. No guarantees are made about what will show up in the new kit, but the rookies at least see the stuff the veteran teams have had the chance to play with. It may be that FIRST already does leave this info up on their website - I haven't checked.

2) FIRST makes available to all at the time of registration AT LEAST the parts list for the goal (not necessarily the whole playing field and stations, etc), so that we can order the fittings, cut the pipe, and so on in advance. That way we can assemble the goal the day after kickoff, or that nite for the local insomniacs. I wouldn't mind also getting the drawing for the goal at registration, but that may not be necessary to essentially take the people-hours of goal building out of the crunch time. It's the time consuming construction - as opposed to simple assembly - as for the puck this year, that I'm not sure how to handle without getting the goal drawing early.

3) FIRST makes available to all at the time of registration the specs and source(s) of the new game objects - balls, tubes, floppies (were they weird enough?). This has been a sticky spot in the crunch time for the three years that we've played. The supplier(s) may be overwhelmed with the sudden rush of orders from all the teams. The surface treatment of the object, like the paint on the Torroid Terror tubes, may not have been fully developed by FIRST ahead of time. And the labor to fabricate objects like (those weird but wonderful) floppies can be substantial. A lot of early crunch time has been spent in the past as the teams fumble around with the game objects, discover the ambiguities in the initial specs (size, or pressure, roundness?), discover the variability in the supposedly 'standard' objects, get answers from FIRST via team updates, etc. I don't want to get too wound up here, but we wasted an incredible amount or time last year on ball manipulators while the ball size/inflation pressure issue was resolved. I think the best way to accelerate the 'object development' process is to get several hundred teams buying multiple objects and then discovering the issues of variability, etc so the specs can get zeroed in. Nobody should, in a perfect world, spend two weeks of crunch developing mechanisms to handle objects demoed at kickoff, and then have to trash that work and start over because some fundamental characteristic of the object gets revised. This applies to play strategy as well as manipulator design, so it can be a pretty big pain. So I'd like to do everything possible to shakedown the play objects in advance.

That's it. I believe that FIRST does an outstanding job in creating the rules of play each year. They miss very little, and they get better each year. The teams identify the ambiguities very quickly, and they get resolved quickly for the most part in the Team Updates. The issues of rule interpretation, they do pretty well with clarifying. Some of them - like what constitutes 'carpet damage' - will always be a bit squirrely, and we need to be prepared. Some - like when is blocking a floppy lift intentionally destructive - can't be precisely resolved by FIRST until we show up and play the game. Bottom line is I don't think we are unduely held up or misled during early crunch time by slow resolution of rule interpretation questions. If that opinion is accepted (a big IF, feel free to disagree), then there isn't a need to issue the rules early and hence define the game. We have the goal parts and the game objects, and the rookies have the old kit list and RC manual. We can drive ourselves crazy trying to guess what the game may look like, but the only jump we really get on design is thinking about how the bot might pick up, throw, squeeze, roll, stack, punt, or otherwise manipulate the new objects.

Anybody who wanted to get a jump on prototyping and developing manipulators would be free to do so, as many of us do already on chassis drive systems, with the accepted risk that the work might turn out irrelevant to the new game. There is no way to prevent early speculative development, and, frankly, this provides an opportunity for some structured skills and knowledge development before crunch time swamps all those good intentions.

Dodd