Everything you thought you knew...

The team met at Mountlake Terrace High School on Saturday morning, along with students from dozens of other teams, to watch the kickoff broadcast of the 2013 FIRST Robotics Competition.  Called "Ultimate Ascent," the 2013 game features frisbees as game objects and two large, jungle-gym-like apparatus in the middle of the field.

An animation describing the challenge can be viewed here -- http://www.youtube.com/watch?v=wa5MGEZNrf0

After leading a brainstorming session at the kickoff site, we headed back to our lab and started our own discussions about how to build a competitive robot for this challenging new game.

Ultimate Ascent is an interesting game -- it features a game object that is unlike anything ever before used in FIRST Robotics -- and couples that with a challenging set of requirements for scoring points by climbing the jungle gym (called the Pyramid).  That means that we're in uncharted water here; there's nothing to look back on when it comes to designing mechanisms to manipulate frisbees.  Teams have built robots to hang from high structures before, but never have they had to do so in such a specific set of steps as the rules for 2013 prescribe. 

First thoughts --

Our first impressions of the game were that it'd quite difficult to throw frisbees accurately and that the point value of climbing to the third tier of the Pyramid -- 30 points -- would far outweigh the value of collecting and scoring frisbees -- even with an accurate shooter.

We considered that a robot carrying six* colored frisbees with it as it climbed might be able to more easily score them in the goal atop the Pyramid, earning another 24 points (after penalties) for a total of 54 points.  54 points seemed, at first, to be a solid contribution to an alliance's score and the team could ignore all other frisbee related functionality (i.e., collecting them, aiming, shooter or dumper design, packaging challenges, etc.).

Other ideas --

 We tossed some other ideas into the arena, as well, although it was pretty clear that the allure of climbing had captured everyone's attention.  One group of students proposed a strategy that focused on collecting frisbees from the human player and scoring them from the just below the goals and another suggested a strategy that required building a robot that could score frisbees in the 2 point goal from across the field at the human player loading station.

Another strategy was to specialize in scoring in the 2 point goal by using a simple arm-based design to raise the frisbees into place and dump them into the goal.  The advantage in this strategy was said to be that it didn't require a finely-tuned frisbee throwing mechanism and could score faster than a robot that needed to align itself accurately with the goals.

When discussing defense, a student suggested we use fans to disrupt the air flow near the goals to send frisbees from our opponents flying off course.

Prototyping begins! --

As expected, the team decided that evaluating the practicality of achieving a 30 point climb was a high priority and about half of us split off into a discussion about what sorts of systems might work for that strategy.  The remaining half went into the lab and split into several smaller groups to begin prototyping frisbee launchers of different sorts.

Even though it appeared as if climbing was the preferred strategy in Ultimate Ascent, we decided it was important to validate our assumptions about the difficulty inherent in manipulating the frisbees.  Unfortunately, I was in the climber discussion and didn't get a chance to video the first tests of our frisbee prototypes, but I've been told that they need 'tweaking' -- one broke itself apart and the other melted a frisbee. :)  Getting those working will be the highest priority for our next meeting.

The discussion about climbers went on for hours and we didn't arrive at any solution that was worth pursuing.  I've spent the better part of the weekend thinking about it -- even waking up in the middle of the night -- and I'm sure I'm not alone.

Next steps  --

Typically, we'd focus on getting an appropriate drive designed and under construction as quickly as possible, but the 2013 rules, in another first, allow teams to decide how big (within limits) they'd like their robot to be.  So, at this stage, even making simple decisions about the size of our drive may be fraught with peril!

Even so, I spent most of yesterday working on a CAD model of a simple, 6WD chassis to get an idea of how much space we'll have available for mechanisms and what kinds of parts we'll want to order.  I've tried to make it relatively easy to change the aspect ratio of the chassis while keeping the design intent of the drive system intact.

This is going to be a tough season, but we're excited for the chance to try so many new things.  Hopefully, we'll have a more well-defined strategy nailed down after another meeting or two and we can begin to build and test prototypes in earnest.  I'll try to get video of whatever we do, though, because this wall of text is probably not very exciting to read. :)

 Until next time,
Madison



 

*We know that robots are only meant to carry four frisbees at a time, but the penalty for carrying more than four is less than the expected value of scoring the additional two in the Pyramid goal, so it's a worthwhile trade.