Every day is exactly the same. We build. We test. We tweak. We build more. It feels like two steps forward, one step back. Thankfully things are starting to go our way. We also got a great reminder about why we're doing this...
"I’m making this up as I go." - JVN
"Aren’t we all?" - Tolly
Day 31: ...One Step Back
It felt rough. We walked out afterward, and no one felt great about the work we'd done. Progress can still feel terrible, even if it is progress. Two steps forward... one step back.
My dear reader... let me help you experience it.
"Close your eyes, pretend you're wearing a very black shirt (the blackest shirt you own). Pretend you're very tired. Irritable. You've been running on the edge of exhaustion for 4 weeks. You're starting to get snappy in your interactions with people. There is less joking and more "head-in-hands" thinking. You have a lot of work to do, but you can't start on any of it because you're waiting for something..."
When we walked into the shop on Day 31, our goal was "test the climber." The climber is the subsystem which has had the least prototyping and testing. The primary goal for the meeting was "test the climber." Everyone walked into the shop expecting to do that immediately.
Then... we struggled all night to get a climber together for a test. Assembly took more time than expected. We realized that we don't have some of the "final" pieces on another mechanism which are needed to test this mechanism. We had to build up some temporary solutions just to get it mounted up for a test. This took longer than expected.
We finally get it ready to go and it's out on the field just as the meeting is winding down. We know that the #2 robot doesn't have a control system on it yet, so we grab an external control board.
Then the pre-test checks begin.
Our team has been doing this for a while, so at least we know these need to happen. We're not silly enough to start the test without going through them. We don't wait until the test fails for a silly reason to go through these checks. We try to find NEW mistakes to make, so we correct these old mistakes before things even start.
Of course, this took longer than expected. We needed to spend time making sure all the motors are spinning in the correct direction and not fighting each other. We don't want the motors to over-heat and burn themselves up. The climber is a 1-way mechanism - it has a ratchet built into it so it only spins one way. So... we needed to make sure the ratchet is set correctly. You wouldn't want to run the mechanism backwards against the locking ratchet... the motors might burn up. Then we noticed the battery is low, so we swap the battery. We're finally ready to test, and... "wait a second, did someone check the ratchet? Did it get flipped during the setup?" Re-check the ratchet. Ready to test in 3....2....1....
"STOP! Hold it... get Hannah. We need to video this."
Getting Hannah takes longer than expected.
Okay, Hannah is present. Let's start this test in 3....2....1....
Everyone turns to stare at the driver Coleman. Coleman sheepishly reaches down to the laptop and enables the robot. Test commences in 3.... 2... 1.........
It failed miserably. It didn't do what we wanted, and it ended up damaging itself in the process.
Silence falls over the group. Some people walk off to work on other things or clean-up the shop. Some people just walk out to their cars, since the meeting is long past it's ending time. The designers are scratching their heads. The designers have their heads down in their hands. "The boss" is looking at me and I can tell she's thinking: "How late are you going to make me stay tonight while you try to solve this?"
I decide. "Let's call it. Everyone go home and get some sleep. We'll attack it with fresh-eyes tomorrow."
Never under-estimate the power of fresh-eyes. Sleep solves a lot of problems.
Big build up to an important test. Failure. Frustration and... back to the drawing board.
Did You Feel It?
My dear reader... Were you bored reading that? Did you sense the frustration of the moment. Did you experience that feeling of "Why can't we just start the test so we can see if this works... because if it doesn't work I really want to start figuring out a solution."
Did you skim ahead? My dear reader, YOU have that luxury. Robowranglers can't skim ahead. We live it in real-time. The process is as important as the ending.
That story, or a story like it happens every night. Multiple times a night. This is just one system getting tested, imagine an entire robot of these systems - always being tweaked and tested and improved. I don't think most people understand how much boredom and frustration goes into the engineering process. You want to understand the Robowranglers and our process? The above narrative is a big part of it.
OF course... we're Robowranglers. We LOVE drawing boards. We eat continuous improvement for breakfast, and our blood runs deep with iteration. We celebrate failures as opportunities for learning.
Ya know what? Sometimes I just get sick of all this celebration. Day 31 was one of those days.
Two steps forward, one step back.
Trust the Process.
After Day 31's meeting, I laid around at home and watched old 148 videos as a reminder to myself that we supposedly know what we're doing.
Blah blah blah trust the process. "You know what... I'll tell you where you can put your process."
This frustration, this anger, this despair, the head-scratching, the boredom, the anticipation, the face-palms, the anti-climactic failure... this is all part of the process. Trust the process. The trick, the only real trick - is to keep going. Two steps forward, one step back. Keep moving forward.
The trick isn't to pretend it is all sunshine and roses, the trick is to accept that it's gonna "suck" and lean into it. You can acknowledge it "sucks", own it, and KEEP MOVING FORWARD.
Anyone who has been through the design of something like one of these robots knows what I mean. They don't come gracefully into this world, they fight you kicking and screaming the whole way.
It reminds me of one of my favorite engineering quotes:
“You can look at practically any part of anything man-made around you and think 'some engineer was frustrated while designing this.' That little human connection makes the object that much more powerful.”
My dear reader... I hope you feel the frustrations, and feel the connection with the fine young women and men of Greenville who are wearing black and like every other FRC student across the world are fighting to do something amazing.
Day 32: Two Steps Forward...
Sleep it off. Wake up. Log into Slack.
The team was already working the problem. Shake it off, work the problem. The #mech-design channel was filled with potential solutions, tweaks and next steps.
Some of them were very simple. That makes me happy - the inclination is to make big changes. To redesign major aspects to fix the problems you see. Sometimes the cause of failure can be more subtle. I like it when people look for the subtle solutions first. We'll try the subtle fix, and then see if we need to do something more drastic.
Set Direction. Execute.
The kids who had robotics class on Day 32 immediately begin modding the robot based on the plan in #mech-design. By the time the meeting started, we're already set to do another test. No "spend the whole meeting getting ready" this time we're off to the races right away. Whatever gotchas Murphy sent at us on Day 31, he benevolently held back from us on Day 32.
A subtle tweak. A big success.
First test of the night. Big win.
Tweak a few things.
Second test of the night. BIGGER win.
Make the test more difficult.
Third test of the night. STILL a BIG win.
Repeat. Repeat. Repeat.
"Is this the same robot we tested last night?"
My dear reader... do you feel the sense of nervous relief? Do you feel like slapping someone on the back and giggling?
About an hour into the meeting, we've already completed climber testing. Everyone is feeling optimistic. The designers are planning their next steps. The #2 robot is passed off for intake iteration.
We have something on the Robowranglers we call "iteration frenzy." Every so often, we'll start rolling down the path of testing something by doing a series of VERY FAST tweak/tests. When we get into these frenzies we can test dozens of iterations in a very short amount of time. This is great for tuning in small variables.
- The external control board was switched over to driving the lift and intake.
- The drivetrain is not wired up. Since Coleman is our base driver, he "drove" the robot by pushing it around by hand.
- Tolly is our operator so she drove the lift and intake using the external control board.
- Using this setup, we scored a LOT of cubes on Day 32.
- The whole time we were scoring, we tweaked variables. Madison designed the Mk5 intake to be easy to tweak and iterate...
- Add some shims to the hard-stop so it is wider even when it is all the way closed.
- OK, now remove all the shims so the opening is smaller.
- Add more spring force to the intake arms so they're held more tightly together.
- Remove ALL the spring force. How does that work?
- Spread the wheels apart more? Shove the wheels closer together? Take off the top set of wheels? Take off the bottom set of wheels?
- Rigidly mount one intake arm so only one side pivots.
- Try adding totally different intake wheels.
- Add... ALL THE WHEELS WE HAVE! Did that work?
- ...No, okay, cool.
- The result of this frenzy? We're feeling pretty good about scoring cubes. By the end of the night we had an intake configuration which we liked. A lot. Time to make it more real...
- The Mk5 was versatile... but now that we have details "finalized" the Mk6 will be optimized for the "final" config.
- "Is this your final intake? It's final... for the moment."
Many FRC intake wheels need to be cleaned. They may "feel" like they're still grippy, but we learned in 2015 that after a little wipe-down with some Rubbing Alcohol they'll perform like brand new. Try it. Maybe your intake performance will change.
In most years, the 148 pit-crew wipes our intake wheels after EVERY MATCH as part of their check-list.
- We're really concerned about the robot weight. The rules say it needs to be less than 120 lbs (not including the bumpers, or battery). All the designers are looking at each other nervously since it is NOT where we want it to be... Time for a robot diet?
- The whole time the above testing / tweaking was going on with the #2 robot, the #1 robot (just a drivetrain and electronics at this point) was going through autonomous testing. The programmers are working through their process. It has the same frustrations. The same boredom. The same failures. The same improvements. The same successes.
- The #2 lift is "done" and sitting on a bench waiting to be mounted to the #3 or #1 drivetrain.
- The next gen climber is fully designed and should be ready for test on Day 36 (Saturday 2/10).
- The #3 controls are done and ready to go onto the #3 robot once it gets a lift.
- The #2 robot is still being powered off an external control board. The #2 control board is ready to be mounted, but we were waiting to install it until the testing was done. During robotics class on Day 33, Day 34, and Day 35 - #2 will be wired up and ready to roll.
"Wait.. does that mean what I think it means?"
- Oh yes, my dear reader... you know it! I mentioned it in my Day 29 update: we're hoping for a FULL scoring system test on Day 35 (Friday 2/9). "Coleman, Tolly... go score some cubes!"
The most important lesson...
Something else happened on Day 32. We all got a little inspiration earlier in the day...
I don't care who you are... that image of those two boosters landing next to each other should give you a little feeling of "ENGINEERING IS AWESOME". Boom. It hit me right in the feels, especially when I remembered this video...
I think about the engineers who work for SpaceX. Some of them FRC alumni.
I think about the Robowrangler students going through Day 31, and all the other "Day 31's" we've had over the years.
I think about how the SpaceX engineers went through their own failures and how they had frustrating days of: two step forward, one step back.
My dear reader...
Think about the way you felt reading the above narrative from 148's Day 31.
Think about how that feeling is an almost universal part of any design process.
Feel the connection with the women and men who made that SpaceX launch possible, how they felt during the times when the rockets were exploding. Imagine the boring and frustrating "Day 31" moments they experienced that you and I never got to see.
Now remember... what lessons the Robowrangler students and FRC students everywhere are really learning from their experience.
SpaceX already put a car in space... what will the Robowrangler students get the chance to do?