Wednesday, May 11, 2016

MATE PA 2016

I joined HTHS Robotics in my freshman year. 
Usually freshmen don't do much. In my year we were just assigned to build props that were never actually used. The upperclassmen were like genius gods who somehow had enough skill to build an amazing robot. HTHS doesn't teach you everything you need to know to build an underwater ROV

In fact it probably only prepares you for 30% of the way. 
Throwing some shade here on VEX, but this is true robotics. Start from scratch, engineer or ratchet your way up to a working robot. We don't believe in pre-built parts, spending excessive amount of money, or outsourcing our design to professional companies.
We believe in autonomy.

2009: The only year we made it to internationals:
According to Zach Darby (2012), the whole robot was built by one guy. Dedication

2013: We reused the 2012 frame, with all the same electronics, attempting to waterproof the pelican box better: failed horribly. The whole thing flooded. I remember the legendary Austin Eng whipping out his pocket knife to cut the electronics cable before it shorted out the topside electronics.
The club sucked then.

2014: Club still wasn't amazing. No one was invested. This year I rebuilt the H-bridge electronics to be on topside. The robot frame was literally PVC duct taped together. To be honest it was wretched. There were still bottomside electronics, I remember. It shorted out of course.

2015: Ayy we dem boiz: The underclassmen become elected and we make a lot of changes to the club. We make all electronics topside and keep the robot design extremely simple. We win Second, qualifying for internationals, but since HTHS is mean, we weren't able to go.

2016: This year. The push year. After the high from second last year we were confident of our major strides from the previous year and aim for first place. This year was so different. Everything was predesigned, engineered the best we could, and precisely machined. It was the most expensive robot our club has ever seen: $1,200. We worked every weekend, and twice a week. 

This year's competition:

Our documentation:

I'm not going to talk about the entire robot in this blog, but I will talk about the week leading up to the competition.
In summary this is what happened
  • Design robot frame
  • gather parts and materials, logistics of robot, planning out system
  • building topside electronics box

  • Still building topside electronics box
  • 3D printing the parts we need, remodeling (first time is never right)
  • Get frame machined

  • Assemble frame
  • Write example/test codes
  • Start assembling parts

April - cram
  • Build bottomside electronics
  • assemble motors
  • waterproof everything
  • Attempt to get a working code
  • build claw

  • Test
  • debug (what I will be talking about)

Our robot is b e a u t i f u l.
Anything more beautiful than our robot is a trillion (exaggeration) more times expensive and had a lot of outside help. Keep in mind that our robot was hand built by a group of high schoolers who were busy with SATs, APs, and HTHS workload on their hands. This is real engineering.

We engineered the shit out of this. I've never seen so much applied science in my life. 

I'm (I was) Electronics President. My primary job is to build the topside and bottomside electronics, and waterpoof the entire thing and make sure the person who is writing the code for the robot doesn't have a hard time. 

This is topside:
You can read about it in the documentation

This is bottomside:
Note TLC and all the H-bridges

note the 5V regulator

In tube

Waterproofed cap

My basement is ruined

This is the finished tube!
Random pics of the rest of the robot:
My basement is ruined, part 2


My car is ruined

ft. the power m8s

So, this is what happened
Sunday - 5/1

  • Test at Pat Meng's House
    • but in reality we were finishing up the code before we could test
  • We found some moisture problems in the tube: A few motors were crapping out, and somehow touching solder joints and measuring wires seemed to fix the problem
  • We have silica gel packets inside the tube to absorb water, but they only absorb if it's in the atmosphere
    • what happens is water leaks in reaally slowly (this time not flooding in, so success)
    • the drips hang onto wherever they leaked in. Over time temperature and pressure difference causes it to evaporate into the tube atmosphere
    • The water eventually condenses on the electronics and the tube walls
    • Things short out and stop working right
  • We added a fan to the inside of the tube. Why?
    • Theory is, if we assist water in evaporating (by agitating the atmosphere), we will get the water into the silica gel packets faster, and therefor water will go into the silica gel instead of on the electronics
  • Did it work?
    • yes, but it backfired.
    • The water evaporated faster, but that just made it easier to condense on electronics. Never again.

Monday - 5/2

  • Meeting at my house to
    • install camera sheaths for fisheye lenses (these ended up not being waterproof and foggin gup)
    • determine Roger's control preferences and finish the code
    • mount cameras

Tuesday - 5/3

  • Meeting at Jaden's Aunts pool
  • Documentation was due this night, we didn't get much testing done.
  • I think we just finished the code, tested it on the table, took pictures, and then we had to go inside
  • We finished mounting all the cameras on. Our distance measurement system (tape measure) was way too heavy
  • The rest of the meeting was attempting to balance the robot in the water
  • Stay up all night to finish documentation

Wednesday - 5/4

  • Stayed home because no sleep
  • Went to Jaden's Aunts pool to finish balancing the robot.
  • Finish Tech report this night

Thursday - 5/5

  • We actually finally got to test. Everything was horrible.
  • The fan backfired, and we only had 1 day to fix this moisture problem
  • I forgot my mother's birthday.
    • this was pretty bad
    • 10k pullup resistors on all H-Bridge inputs for electrical stability and durability
    • Tape silica gel packets on strategically determined positions to assist in moisture absorption (AKA put them where the leak is)
    • Spray the whole circuit board in RustOleum NeverWet spray (smells horrible)
  • I got no sleep.

Friday - 5/6

  • Reupload code
  • Close the tube
  • Test
This was the first year our robot actually functioned.
  • Motors were fast enough
  • Robot was balanced
  • Our claw could actually pick something up
  • We could easily get 100/260 points
We stayed up that Friday to do our ritual preparation over Google Hangouts
  • Optimized task order (to get most points)
  • organized ride logistics and t-shirt distribution
  • The claw broke for the 5th time that week so I had to stay up again to fix it.

Saturday - 5/7

  • Ray and I wake up at 5 to drive to Villanova. I was surviving on 6 hours of sleep for the past 2 days and 2 cans of Monster.
  • The tube had moisture but we weren't worried
    • At the pit we realized that our motors were crapping out again. We used a hairdryer to fix that problem and put it all back together.
    • Our robot worked perfectly. We could last 15 minutes in the pool without having a moisture problem.
  • Get to the pool
    • wait an HOUR because MATE PA logistics are the worst.
    • Plug in the robot
This is when everything falls apart.
We usually set our power supply to 12.3 volts. When we plugged in the robot our voltmeter read 12.7
  • Run the code
  • Nothing works at all. except for cameras and servos, but no motor works.
    • We panic.
    • we collect ourselves, check all connections, check the code, restart everything
    • our 5 minutes setup is up. Now we cut into our demonstration time.
    • I thought that perhaps we tightened the tube so much that a cable pushed the Arduino Mega out of it's sockets. We open up the tube so I can perhaps smear my hands on the electronics and hope something works.
    • Nothing works.
    • We close it up and throw the robot in the pool to at least measure temperature and use the cameras
      • The cameras are all fogged
      • The code crashes from attempting to get serial data from a killed Arduino.
What I suspect happened is the unexpected large power supply killed our TLC or our Arduino, and none of the motors worked. Why did the servos work? I used an external regulator that could handle up to 20V for those motors.

This was supposed to be the year. But we just got really unlucky. Every year I see people jeering at us for having such a low score. This year it seemed like everyone was sulking in the pit, trying to debug their electronics. No one from lane 3 (our assigned lane) seemed very happy about their performance.

One thing in common? Power supply. I'm not saying the power supply was wrong, but there is a good amount of implications. We still need to autopsy our robot. RIP $1,200.

After that we were shoved into a presentation room of hungry and impatient judges who cut our presentation off mid sentence. I have to be honest, those were the most rude judges I have ever seen. I expect 20 year old college kids to have more decency and patience to accept handouts and allow exposition without letting their stomachs affect their mood...

Below are pictures of everything

Our poster that somehow only won 3rd :(

The team that cheered us on and kept our spirits high

Waiting for a goddamn hour


We are ded


Robot in the car

Robot that was alive on Friday

Carrying everything to competition pool

Realizing nothing works

Trying to check all connections

All the tasks that we didn't complete

We're done

In the pit, crying

Somehow we won 4th out of 26 teams? With a robot that didn't even swim?

Though we didn't win or even place this year, I'm still extremely happy with the performance of the club. We got more underclassmen involved than ever, and achieved our ultimate goal to finally build a working robot that looks decent. 
Sadly Ms. G is trading the club off next year. But I'm glad that the club members were able to get involved, and that next year they will NOT have to build from scratch like we have done every year.
Ultimately, 2017 is so set up for first place from 2016. You're welcome. 
I hope you guys win so hard next year that N.E.R.D.S., Excelsior, SEA RAMS, and all those teams get so rekt that their sponsors all flock to us.

If you want more info about the ROV or what we did, you can comment and I'll probably get an email.