Day 11's meeting felt like the last calm before the storm. During our discussion tonight we kept bumping into trade-off driven questions. Soon we'll need to choose one of the paths in front of us... Decisions are coming!
Team Update Balance
Big news today from FIRST in the form of "Team Update 3". In addition to the Rule Q&A system, FIRST periodically publishes small tweaks to the wording of the game rules in the form of team updates. Today's update included a change to a rule regarding the "randomization" of the field. Instead of all the scales being randomly assigned their Red/Blue sides the new setup only randomly selects one of four configurations. These configurations are each symmetrical, so neither Red or Blue has an advantage in any given field configuration. (If my 3 goals are all on the left side, their 3 goals will all be on the right side -- even-steven.)
I think this is a FANTASTIC move on FIRST's part. That unbalanced field would have randomly (literally chosen by a Random Number Generator) given one alliance an advantage in critical matches. Bad enough to be out-gunned, but it doesn't sound like fun to also find yourself playing with the field tilted against you.
While many in the community are quick to point out "life isn't fair, the game still isn't fair, and randomness still plays some role in the competition" - I'm quick to point out: "I'm personally glad that randomness doesn't play a role in this particular aspect of the competition." Thank you FIRST!
Today's Robowrangler Huddle was a LOT of fun for me (and hopefully everyone else) and I think the discussions went very well. A big part of the reason it was such a success, is that I brought in an amazing outside resource...
Today I wanted to talk about lifting subsystems. We've done a lot of work on intakes to pickup and hold the cubes. We've done a lot of work on climbing mechanisms to lift ourselves up onto the bar in the endgame. We've started working on drivetrain design. We've even played with the way the cubes will be scored... but we haven't talked a lot about lifting those cubes off the floor to the height of the goals.
There are lots of presentations out there talking about the different types of lift mechanisms. There is a famous one which has over the years been passed around the community like some sort of rental car. I've written some presentations of my own... I've never been happy with the way I teach this subject (even though I literally wrote a book on the subject).
Ladies and Gentlemen... we can do better than BAD powerpoint.
Every so often you encounter an educational tool which totally changes your perspective on a topic. A master class. An educator at the top of their game, presenting the material in a way which hits all the key points, but still makes it accessible to their audience.
A GENRE DEFINING resource...
For Your Consideration: Aaron Palardy, Robot Expert.
Note: I'm not kidding. I LOVE this video. I think he does a great job of showcasing different types of lifts, and visually demonstrating some linkages which newer robotics students can't separate from the jargon. (Have you ever tried to explain a double-reverse-four-bar to someone without a video or physical model? Try it. Do it with just a piece of paper and a pencil, I dare you.)
The Killer Bees will be lucky to have him someday. Unless... he wants to put on a stylish black tshirt instead?
Aaron, if you're reading this: "You don't need to settle for the team which coincidentally builds near where you live. Shop around. You could be a Robowrangler!"
Bryan's design group talked through the work their team has been doing with lift concepts. They did a variety of 2-D mock-ups which experiment with lift packaging and test concept reach.
If your team is designing a lift subsystem:
- How tall does it need to go? Are you scoring on the scale? When it is tilted away from you so your side is higher? Do you want to put a cube on top of another cube already on the scale? Do you want to be able to put a 3rd cube on top of a stack of 2 cubes already on the scale?
- How low does it need to fold up? Do you want to be able to drive UNDER the scale? That means the robot height limit of 55" is actually too tall. I think you need to fit under like 42" to safely drive under the "low" config of the scale. Does that matter to you?
- Mini-Palardy talks about some concepts having "sweep problems." The robot can never extend more than 16" outside it's starting frame perimeter. Does your lift concept extend out further than 16"? How do you package it into your robot?
- How does your lift design affect the height of your Center of Gravity? Lower CG is always better. It affects robot performance in areas I never expected.
- What about front/back CG? If you're lifting some weight up "high" that will make the robot more tippy. When you're "up" you need to make sure your CG stays firmly over the top of your support polygon, even when you're moving around. (Translation: don't fall over because the robot becomes too back-heavy and you drive forward too fast.)
- How does the lift package in the width of the robot? How does it integrate with your end-effector? Where is the cube "hanging out" as the lift is going through it's motion. Some lifts will be better for wide-body robots than long-body robots.
- Does your lift need to change the orientation of your "claw" as you lift it? Maybe if the lift automatically tilts the claw as it rises, you can get more height out of it.
- How much value / benefit do you place on "Sweet linkages". I place high value on this. I'd rather have a "sweet linkage" than a simple and effective design. The Robowranglers probably shouldn't let me give input on our final lift decision.
Trade-offs & No Clear Winner
As the discussions around lift subsystems came to a head, and as the different prototyping teams reported back in it became apparent we have a problem. There doesn't seem to be a clear "winner" for the overall robot plan...
We have some competing philosophies which are all still "in play" for this year's design. We have different aspects of gameplay and strategy which all "seem reasonable". We have different mechanism concepts and prototypes built which all seem like they could perform well.
There is no CLEAR winner, and there are pros / cons to each option. We can keep prototyping, but I don't think we're going to learn enough to make our path clear.
I don't think we can build 2 different full robots each with a different strategic philosophy just so we can see which one works best... but it's not like we haven't talked about it. We're not that crazy or desperate (yet).
I think we will discuss the options as a team, and then we'll just need to make a decision and set direction. I expect this will happen Friday or Saturday (Day 14 or 15).
What will we do until then?
Keep exploring, keep detailing, and keep learning!
Things will move fast later, if we do the leg work now.
Corbin tells us his Mom reads the blog and asks him about his thoughts on different things she's read. ("Hi Corbin's Mom!"). He says this is exhausting since he's tired from a long day of robotics. We think he's silly, and should be happy to talk robots with anyone, especially his Mother.
Message to Corbin's Mom: "During Tuesday's meeting on Day 11 I didn't work with Corbin, but I know he worked on CADding the drivetrain with Jessi. They were playing with different wheel configurations based on that thing we realized yesterday (and I mentioned mysteriously in the Day 10 blog) which might make us re-think a lot of our design. You should ask him about it!"