Week 8 - Determination

So much is together, but still more work is needed.

Our 8th and final week of build season ended slightly earlier than the others, as Friday at 5pm Pacific time, the Glacier Peak competition started. This was the final stretch to prepare, test, and tune as much as we could. Also, we still need a new intake. I'm mentor Bobby Neckorcuk, and this is the longest week of my life. <Intro music> (Does anyone recognize that paraphrased quote anymore? That seems really old now...)

This intake. I feel I'm going to just keep talking about it. On the plus side, one strength of this team is how we improve our knowledge and process with every iteration. We normally don't meet on Sunday's but a well-timed Discord message led to us getting connected with another local team. A lunch-time CAD review, one team mom driving a sheet of polycarb to our house, and an hour drive later, our good friends on team 1983 were preparing our 7th set of intake plates. By Tuesday, we had our newest intake assembled and mounted on the robot, ready for our next session at the practice field. We were in business!

Huge thank you to our friends at 1983 for cutting these plates!

Work continued on our final subsystem - the climber. We had both towers assembled, but were delayed working through some friction issues. What did the internet say? Simplify - and add grease. (Lots of grease.) Feels weird to say, but we were on the final assembly of the final subsystem of the robot - the climber winches. With our nice simple "climber in a box" solution, we opted for the same "winch in a box" to solve this part of the problem. For some space saving and packaging, we had a plan to mount one bracket to support the climber winch, but after seeing an Open Alliance thread where a team had a similar solution then bent the bracket and failed a climb, we did a little shift and decided it would be in our best interest to mount both plates and fully support the winch drum.

My bet is he's looking for more grease.

Our burndown list became tiny. Sponsor stickers were soon the highest priority item on the list (for mechanical anyway, software always has more stuff to do. Sorry software!) We installed a chain tensioner - a 3D printed part with an open slot that mounts on top of the planetary gearbox. The 3D print was super nice, as we didn't need to perform any specific machining for slots or curves. One final electrical item is zip-tying all our Anderson connectors together. That is easier said than done on this tiny robot, I hope we didn't miss any! (No foreshadowing is happening here. Move along.)

I have definitely lost track of how many intakes we've built this year.

Software always has work to do. Our bring-up process was all about slow, predictable, measureable movements. (Don't run 100% power the first time you spin a motor!) Now that we felt we had some level of command over the controls, we could start speeding things up. Our 'stow' to 'speaker launch' position was just about 3 seconds, and we dropped it to around 1 second. We doubled the speed of our wrist rotation. We kept our climber motors at 20% power, cause, well, they weren't there yet.

Slow motion.
It's faster now.

One final challenge for software was handling the Note inside the intake itself. For reasons, we hadn't added any sensors to the robot (besides the built-in motor encoders). Our Note intaking and automatic stow was based on reading changes in amperage usage. Our plan for launching was pretty much similar - detect the drop in amperage when the front rollers are no longer trying to move the Note, then spin to launch rpm, and the rear rollers will feed in the Note forward. Theory is a great place, but we still hadn't actually implemented and tested our planned code. Once again, patience and Shuffleboard helped us out tremendously. We gathered data on the amperage without a Note, with a Note, rinsed and repeated <many> times, then built thresholds in code based off measured values. (I wonder if we should re-check these before every event...) It turns out theory is a nice place, and the solution worked! Quite beautifully actually. Below might be looping gifs, but during our shop testing, it really was that consistent.

Om nom nom nom

nom nom nom

We had practice field time on Tuesday and Thursday as our last chances to 'run what we brung' and see how the robot will perform in earnest. We ran our basic auto paths for the first time, and sorta re-wrote almost the entire operator control command scheme. Tuesday we had a fun issue on our first runs. We bumped into the Stage, and lost radio. It came back, we drove into the loading zone, and again lost radio. Last year I watched 2910 do this one weird trick that robots hate - drop the robot 3-4 inches off the ground and stare at any and all status lights you can. My turn to throw the robot around! :-) Low and behold, after two drops, the loose wire in the Roborio came free, and showed us why we were having intermittent issues. With that more securely attached, we ran mechanically and electrically faultlessly for the rest of the evening. (Do you believe me now when I said earlier that there was no foreshadowing?) 

I'm very happy to find these at the practice field,
rather than after a competition match!

Our intake, out beyond our wrist, at the end of our arm, wobbled. (Technical term.) We got this feedback way back in week 3 that we will really need to watch our backlash, because it will compound through the two rotational joints. If nothing else, we can at least add bias the backlash, so we added another constant force spring to pull on the intake, providing some stability so we could launch and score with more consistency. By the end of Thursday we were able to ground intake, cycle, score in the amp or the speaker, and our operator and driver were starting to build some of that muscle memory to perform actions quickly together on the field.

Friday, our final hours of build season were all about the climber. We installed the first tower, winch, and tension rods. We manufactured the next set of tension rods and got the winch assembled. The minutes counted down to 5pm. One final time the robot was handed to software for a systems check - run through all the functions and verify everything is still working as expected.

The main breaker was switched off.

The robot left the shop.

Competition had begun.

Quote of the Week:

Mentor 2: "Oh right, left hand taps go the other way!"