There are a lot of common events that FRC teams will go through over the course of build season. The general schedule teams see is: Hype! Intense Focus! Great Ideas! Some Building! and then... the waiting. And some more waiting. And then OH-MY-GOODNESS-ALL-THE-PARTS-ARE-HERE-AND-THE-MACHINES-ARE-RUNNING-AND-HOW-ARE-WE-EVER-GOING-TO-BUILD-THIS-THING, and then the drive team and software kick into high gear while everyone else... waits... Right now, we are still in the waiting phase. As we machine all our parts in house, we have a few very nervous students and mentors who are looking at the flood of parts that will soon be let loose upon them.
Lessons Learned
On a high level, our team learns valuable lessons every year. Internally we see what mechanisms worked well, which ones could have used another iteration. Externally we see hundreds of other robots and how they solved similar challenges. Teams in the Pacific Northwest and across the world host workshops on software, design, fabrication, strategic analysis, and robot reviews. Many can be found online, or linked from team websites. Across all these tools and resources, we have come to learn a number of very important lessons, some of which I will attempt to describe below!
* Use the tools provided! Teams share out CAD, software, motor/spring calculators, and more. Learn how to use them and force yourself to make use of them during build season. During the off-season, understand how they work and how the calculations are made, but during the season, time is a diminishing resource, so use the tools at your disposal.
* Build what you can. Don't build what you cannot. Are you a veteran team with lots of resources for fast iteration? Are you a rookie with younger, less experienced students? The two team descriptions should have different team goals, and different robot designs. Build what is within your capabilities. (I'll let you know at the end of this season if we ignored this lesson....)
* Days one and two are the most important days of build season. Energy is high, minds are fresh, you are seeing the game for the first time. The observations and decisions made on the first weekend will impact every day following. If you have a solid productive weekend with good decisions, it will positively impact the entire rest of your season.
* Review everything! Did you finish a piece of CAD? Did you finish your overall robot strategy? Take a step back, have key stakeholders (engineering mentors, drive captains, etc) review the decisions made and why inevitable trade-offs were made. Reviews are used to strengthen and improve the final product. Review all the things!
* Give time for software, and give time for your drivers. One, they will be happy, and happy software is good software. (1's and 0's have feelings too!) Two, your robot will perform better. Our first year, our software team had roughly 15-20 minutes of robot access before stop build day. First day of competition was a nightmare of testing and debugging. In 2016 the first time our drivers actually drove the robot was in the practice matches at our first event. In 2015 we produced arguably our least capable robot, and yet with plenty of software time and drive practice it performed at a higher level than the robot should have been able to achieve.
* Build control-able systems! The fastest assembly in the world won't score if software or your drivers are unable to use it. Robot drivetrain speeds should be considered for the given years field. Gearing and motor decisions should be made alongside sensor and feedback options. The best possible end of this is having your software team able to read or set any position of a rotational or linear system. Plan for sensors, and plan for designs that can be operated and controlled.
* Speed is king. Cycle time is king. Point densities are king. Touch it - own it. Everything in FRC is fast. 2019 especially, one cycle can be the difference between a win and a loss. (Across FRC history, semi-finals, finals, Einstein matches and more, many games have been won by 10 points or fewer.) Reducing the margin of error on your intake and deploy mechanisms will enable faster cycle times. Avoid dropping game pieces at all costs. Be able to pick up any game piece in any position to short-cut cycles when possible. (Another lesson we may be ignoring this year. I should really have written this post sooner...)
* Secure all of your electronics. Stronghold taught us many lessons. The high level of this is "Play every second of every match". Secure electronics won't come loose, won't drop batteries onto the field, won't cause brown-outs or restart the RoboRio. Put the same time and effort as your mechanical design into your electrical (and pneumatic) design.
* Self-inspect your robot. Day one of every event everywhere is for load-in and inspection. I don't personally know of any event where *every* team has passed inspection on the first time around. Definitely don't be the last team to pass inspection. Start early. I personally print out the inspection sheet as soon as it becomes available during the year, and as we are prototyping, CADing, and assembling, I am constantly comparing our robot features to the inspection list. When our robot is "finished", it goes through another inspection. Before the robot goes in the bag, we give it another inspection. You should never show up to an event unprepared for inspection.
* Your business and marketing team is SUPER important. You want to build a robot? Want materials? Want electronics? It all gets paid for by someone, from somewhere. Sponsors understand and see value in supporting STEM programs, and FRC has the advantage of being a "physical product" program. Want to impress a sponsor? Bring your robot to them and show off what you did, what you learned, and what you would be able to accomplish with more resources. How do you find sponsors, or how do you let sponsors find you? That's all the marketing and social media aspect of the team. One of our largest donors found our team after seeing an article in a local newspaper. Our team still sends regular updates to this this donor, and all of our sponsors. (This is a great time to plug the Compass Alliance. Their resources for submitting awards, fundraising, and sponsorship are the combined product of many world class programs. Check out www.thecompassalliance.org for more!)
* Apply lessons from school toward FRC. Apply lessons from FRC to life. The goal of FIRST is to change the culture of the next generation of students and graduates. FIRST teaches how to analyze a problem, test different options and solutions, continue to iterate and improve, and continue to learn. It teaches us how to simply get out there and make something. These are not simply lessons for your next test or exam. (Unless there's a new class that I didn't have at my school..?) They are how myself and so many mentors approach their career projects. It's how entrepreneurs operate. It's how the economy of the US operates - creating, building, and learning continuously. (You know, when the US economy operates...)
Oops
On the topic of lessons: numbers, distances and extrusions are very important. Imagine a world where a robot deck is completely flat. Now imagine a critical part, like, a battery, protrudes 1/2" above that deck-lid. I wonder if that impacts any part of the design that has been so well integrated and volume-d. Oops. That's not gone well. We had a number of systems and discussions about volumes, margins, mounting/interconnection plates and more, and yet, today was the first day where we added every subsystem into one giant model. (Folks said we were daft to open Solidworks on a laptop. First model crashed. So we opened a second one! That also crashed. The third model exploded, caught the RAM on fire, deep-fried the GPU and then crashed. But the fourth one loaded up!) This was a great lesson for a few mentors and students alike. For mentors, we make assumptions based on dealing with students from prior years, however, each year is different. For students, sometimes they don't know what they don't know. So when they talk about where a gearbox mounts, or what clearance they have with another system, (or how flat the deck is), they take a verbal answer at face value and continue moving forward with those assumptions in place. For all parties involved, we need to keep in mind the opposite extremes. When finishing the details of this model, we will check the micro level points such as - will a rivet (and tool) fit in this corner to provide the necessary rigid structure. On the opposite end, we need to remind ourselves that each rivet, each sheet, and wheel, and bearing, is part of a larger whole. Every now and again we should step back and take a look at the robot as a whole. (If you do your reviews on a mountaintop ledge, please look behind you before taking a step back.)
Head Coach quote of the day:
I'm reaaaally trying hard not to panic right now - Mike
No comments:
Post a Comment