Design Checklist?

[Durgecaller]

I was assigned with the task of drawing up a robot design checklist for my team's brainstorming period. The basic function of this list is to not only encourage basic design ideas, but it will also function in winnowing out the designs that will be brought forward within the first few days following Kick-off. I was wondering if there were any suggestions that the really experienced Vets had for a second year team on such a process. I have all of the basics: Drives, Payload Handler, Basic cost, overall functions... but they just don't seem like enough. The plan is that the team will split into evenly numbered sub-teams after the kick-off and begin to brainstorm ideas of designs for this build season's robot. I know that it is possible to create a list without knowing what the objective is for the 2010 build season, but I have hit a mental block and am looking for any suggestions.

[rsisk]

Anyone serious about the design process should review JVN's document from this thread: http://www.chiefdelphi.com/forums/sh...&highlight=jvn

[David Brinza]

Originally Posted by rsisk

Anyone serious about the design process should review JVN's document from this thread: http://www.chiefdelphi.com/forums/sh...&highlight=jvn

This is an excellent resource.

Remember that it's better to invest time upfront deciding what is the right robot to build, rather than discovering later that your robot concept is complex. difficult to build and ineffective. Strive for simplicity in your robot design: fewer things to build and fewer things to fail is a good philosophy.

[IndySam]

Andy Baker has a nice presentation.

Designing-a-FIRST-robot.ppt

The matrix they use is a great place to start.

[Clinton Bolinger]

I would recommend that your teams strives to understand the strategy of the gmae before even starting your design phase (talk about drive, payload handler, etc).

Understand What you are going to do before you determine How you are going to do that.

This also helps the design phase by allowing you to have design constraints to refer to when brainstorming/prototyping.

-Oris-

[sanddrag]

Don't write off space and mass for the electronics (including wires), and a good solid battery mount. Also, integrate bumpers into your design (assuming they are required again.) And for everything you design, ask yourself if there is an easier or better way. Finally, when you design something, ask yourself "WHO is going to make this, WHEN are they going to do it, and HOW LONG will it take?" Many of these things seem to be afterthoughts for many teams.

[TEntwistle]

JVN also has a great video lecture available through the FIRST website that details the design process. It is an hour long, but a really good look at the process from beginning to end (with plenty of humor mixed in). One important message of the video is to determine your need first, then how you want to accomplish it, then how to build the robot to make it there (along with several other steps along the way). For example, if your team decided that it needed to score 20 points per game under Lunacy, you could figure out how you wanted to score them - robot scoring moon rocks, human player scoring moon rocks, or super cells. After you have that decided, you make your robot to achieve that specific goal. You can make all of the checklists you want to now, but until you know the game, and then your scoring strategy, you won't be able to get very far.

[Ian Curtis]

Originally Posted by sanddrag

... and HOW LONG will it take?" ...

Then double that number to figure out how it will actually take.

Originally Posted by TEntwistle

For example, if your team decided that it needed to score 20 points per game under Lunacy

Important point: You'll usually overestimate the number of points needed (2009 was a rare exception due to the HP element, IMHO). In 2008 if you could average 2 hurdles a match, you were a first round pick at most events. In 2007 if you could average 3 tubes a match, you were probably a first round pick. In 2006 if you could make most of your starting 10 balls in every match (or score well in autonomous), you were a first round pick.

Also, is there a link to JVN's video lecture?

[TEntwistle]

JVN's lecture

http://thinktank.wpi.edu/article/156

Definitely worth a view

[Chris is me]

Originally Posted by iCurtis

Important point: You'll usually overestimate the number of points needed (2009 was a rare exception due to the HP element, IMHO). In 2008 if you could average 2 hurdles a match, you were a first round pick at most events. In 2007 if you could average 3 tubes a match, you were probably a first round pick. In 2006 if you could make most of your starting 10 balls in every match (or score well in autonomous), you were a first round pick.

I'd say it depends on your goal. If you're trying to be an alliance captain, trying to be an alliance captain at the Championship, or trying to WIN the Championship, those numbers need some adjustin'.

[JVN]

Originally Posted by TEntwistle

JVN's lecture

http://thinktank.wpi.edu/article/156

Definitely worth a view

I'm glad you enjoyed my presentation, and I'm happy you got something out of it.

Thanks for posting a link to this. I had not seen this video before.
Having watched it, I don't think I'll ever be able to give the presentation that well again. If they invite me back next year I may just show this video.

-John

[Ian Curtis]

Originally Posted by Chris is me

I'd say it depends on your goal. If you're trying to be an alliance captain, trying to be an alliance captain at the Championship, or trying to WIN the Championship, those numbers need some adjustin'.

IMHO, It's the majority of teams that shoot to "WIN the Championship" that need to realize 2 hurdles would be fantastic. 2 consistent hurdles was probably within striking distance of a good alliance at most events in 2008 (and at the others, with a bit discretion would make you a *great* second round steal). For some of us, winning a regional is a prereq for making it to ATL.