Day 11, 12, 13

On this day last year, we had a wired rolling chassis. This year, we don't.

Guess which team from whom I stole that picture. :-D

Deadlines Appear

We are starting to get into deadline mode! Chassis must be finished. Anything that touches chassis must be integrated, the mounting plan finished. Mechanisms themselves are moving from prototype to CAD to review to machining. Silver robot must be in software's hands in fewer than 3 weeks. Deadline all the things! Integration is taking center stage now as teams vie for deck space on the chassis. Each team uses some real estate, either as superstructure that is deeply embedded in the chassis or fixtures requiring strengthened mounting brackets. Electrical needs a surprising amount of space to properly mount components, as well as considerations for cable runs and pneumatics. A clean (labeled, documented) robot is an easy-to-work-on robot. Schedule is our number one robot goal, deadlines are our on-going measure of success.

Building Prototypes

Our first physical prototype this year was built by our software team. To be fair, our mechanical team has been using a ton of CAD prototypes and stealing mechanical prototypes from previous robots. Usually, we start building prototypes out of wood. Chop a few pieces, drill a few holes, stick a PVC or round axle through, and stick a drill on the end of the shaft. Boom, we have something that can spin. One can cut longer or shorter wood arms to test (or verify) geometry around a game piece. 

Software built a prototype to test viable options for our LimeLight. New for this year - we are guaranteeing solid mounts for all of our sensors and feedback, included in CAD. We started this process last year, but did not utilize it in full. We had LimeLights ordered and tested, but we were prototyping and match drilling mounts directly onto our practice robot. This year, our prototype will validate available positions and view angles, and that will get fed into our design teams with mounting holes added directly into CAD. Our electrical team took the cardboard approach. Cut out sections of cardboard, place or tape them together, and draw outlines around each component where it will mount. They have a defined volume from CAD, so as long as their cardboard shapes fit within that volume, they can use cardboard prototypes to plan ahead.

Our prototyping follows two main goals: we prototype for packaging/integration, and we prototype proofs of concept. The best prototypes are the ones we don't have to build. Robot in 3 Days often show geometry or ideas that conceptually work. Taking solutions from previous years robots also usually do not need much physical prototyping. Generally, physical prototypes are built for mechanisms that directly interact with the field and/or game elements. This year, we have built and borrowed physical prototypes to see how different materials and intake setups handled the cargo and hatch panels. Our prototypes and CAD models evolve in parallel. A prototype may show a new geometry or arm length better meeting our robot requirements, which would get fed into CAD. CAD would then check packaging and integration, and feed any information back to the prototyping teams.

The Opposite EveryBot

I like thinking about things. During build season, I often think about specialized bots that don't fit with our teams robot requirements, but just do something... different. This year, I thought about point densities. I actually thought about the same last year. Sandstorm and end-game have the highest point densities. What if a team put all their resources into climbing, and the only thing their robot did was get to the 3rd level?

* 12/154 points

Well, less than 10% of all available points with one action. We can do better. What if this team now brought a friend, or 2 friends up to level 3 with them?

* 24/154 points
* 36/154 points

Woah. one action, almost 25% of a maximum possible score. And we just scored an RP. That's pretty cool. Thats also just in end-game. Let's add a simple sandstorm - if we climb down from level 1, better yet level 2 and cross the Hab line!

* 42/154 points.

27% maximum score. All by yourself (ish. You need your partners to be able to drive onto a 30" platform). With nothing else, 42 points will win quite a few district matches. But, maybe we still haven't spent all of our engineering effort yet. Lets slap a single strip of velc - er - hook and loop material somewhere on the bot, and stick a camera to sight the ... hatch panel location. New sandstorm - drive off level 2, use the camera to drive over to cargo ship and stick a hatch panel. In endgame, climb up to level 3 with your best friends.

* 47/154 points.

5 extra points (2 for hatch, 3 for pre-loaded cargo) puts us over 30% of a maximum possible score. Plus an extra RP. With 2 real actions performed. And well, 2 minutes of your drive team sitting around twiddling their thumbs, high-fiving, jumping rope, hastagvictoryspins, whatever bored drive teams do. How cool does that sound? While I definitely do not recommend a season start with this type of gameplay analysis and evaluation, it is interesting to think about. I do hope some team has thrown their resources toward a solution such as this. It is not without difficulty, lifting 450 pounds of robot over 19" in some form or another with these game restrictions will take some solid engineering. That being said, we will see it happen, and I will smile.

Yet another commonly repeated phrase in the FIRST arena is about how every team will solve the same problem differently. Team goals are different, robot goals are different, engineering support and capacity is different, and this leads to thousands of different concepts, ideas, and implementations. I love seeing super well-thought-out efficient designs that can play the game at a high level. I also love seeing unique and innovative solutions brought about by limited time, resources, or materials. Think about things. Think differently.