Thursday, September 5, 2019

Reflections from 4 years of Silent Spring

To be honest, I didn't follow the Combat Robot Hall of Fame so I had no idea that folks had nominated and voted in Silent Spring until it already happened. Still, I recognize this is a huge honor! It is recognition from my peers and competitors that Silent Spring is a bot to be remembered. In addition, it is the first beetleweight to be inducted as well! Wow!



I took this chance to update the page on Silent Spring (especially the picture >_>) and realized how far the bot has come from the initial design. I first up-scaled DDT just to have a beetle to compete alongside my ant. Dominant Mode was totaled the year prior by Aaron and Margin of Safety so I needed another robot to spend my time at events. At first the bot sucked; it overheated on the weapon frequently from being undersized and the drive was tiny so it couldn't maneuver or dodge opponents. Still, when it connected a solid hit it made such an explosion it made me want to maintain it. Bit by bit I improved its reliability until, despite the huge impacts weapon-to-weapon, it survived and started winning.

It is because of this evolution, I had a chance to reflect on how far my own engineering has come. I will share here, briefly, 10 things in random order I have learned that all builders could benefit from.

1) If it is supposed to move, leave clearance around it... and then add more clearance.
One mistake I see many builders make is they design their bots too compact and tight. Sure it looks nice, but you have to remember that metal (and plastic) can/will bend and your bot will take damage; if it bends and jams up your drive motor or spinning weapon, it is as good as done! Adding some extra space ensures you can take some abuse and keep going. This goes for wires for those pesky motor-in-drum weapons too.

2) Don't over constrain your bearings!
Overconstraint is when an object is held firmly by too many ways simultaneously. This may be fine with structures, but for bearings and mechanisms its a huge efficiency loss. For bearings specifically, it preloads the bearing which can add friction, decreasing max speed and increasing failure rate.

3) Test ahead of time (but do it safely)!
In addition to competition preparedness, you also need to make sure your bot will work reliably. This doesn't mean build the entire thing and smash stuff on your drive way, it also means testing out and verifying performance on a component-level. Especially with new technologies or mechanisms, leave yourself time to understand them fully and find out if they will fail in your application. I remember the first time I tried brushless drive on Attrition, I ultimately failed because I was under a tight timeline. There is a lot you can test before putting the bot into the arena!

4) Be prepared!
The number of competitor robots at an event is at an all-time high so matches will come quickly. Sometimes you win and sometimes you lose but in either case you must be able to recover quickly otherwise you will lose from bleeding wounds. Make, purchase, and prepare your spares. Assemble them into nearly battle-ready states before the event. Anticipate where you will take damage and make more spares of those. Being fully prepared may take hours or even days but in contrast that is time you wont have at the event. Better safe than sorry!

5) Don't underestimate the importance of padding around the delicate stuff!
In this mechanical sport, its easy to forget about the electrical bits and they aren't nearly as durable. I have had electrical components fall off the board from impacts and it is both hard to diagnose and incredibly frustrating. However there is an easy solution: just add padding! Padding smooths the motion of components held within so they don't feel as much shock. Make sure to remove hard connections to the electronics so they can move. Also be mindful of the wires! Strain relive them so the electronic component's inertia doesn't pull the joints apart and protect them with sleeving to they don't abrade on internal surfaces over time.

6) Just because it is squish, doesn't mean it cant be armor.
A large focus on armor and defense has been hard armor and deflecting surfaces. It is a good strategy but hard armor in most cases means metal and that is weight that many bots cannot afford. An underappreciated defense strategy has been using compliant mechanisms like rubber or thin plastic sheeting to absorb energy. The idea that a material in a certain geometry can deflect within the elastic regime so it doesn't deform permanently and spring back. Combined with a naturally tough material, this can be very good for surviving glancing hits or blunt impacts. Just ask HUGE!

7) When in doubt, make it stout.
Just like the first version of Silent Spring was very glass cannon, if you want to win reliability is your friend. It is okay if you aren't the strongest, being tougher will get your further. So oversize the critical parts just a bit more and outlast your opponent.

8) Do not use screws in shear!
Screws are made for tension, not shear. Shoulder bolts are okay because the shoulder is hardened and acts as a pin.

9) Build something crazy occasionally: try something new!
I love maintaining my competitive designs and trying to innovate upon them bit by bit but it is super fun to try a completely off-the-wall design. Some of the most fun moments of my robot fighting career have been making silly robots (assbots) with my friends. Some examples of bots I made with my friends include "prop quiz", "melty butt", "HAM slicer", "exitbot", and the OG "colsonbot". More recently I built "green eggs and HAM" with Xo and Aaron. Think of it as a dumb melty brain spinner x SOW. We built it quickly out of garbage but miraculously it worked and showed some potential for a future bot type.

10) Have fun!
Because if you're not enjoying what you do, what is the point?!