Motivations
Wherein I describe how I came to robotics as a hobby.
In late 2016, a short blurb appeared in a list of upcoming events at a makerspace near me called Artisan’s Asylum. It read:
MASSdestruction Robot Competition
Saturday November 12th
10AM — 6PM
The home for robotic combat in Somerville Massachusetts! Battling in the Cochran Combat Corral, see some of your favorite local robots battle each other in a variety of weight classes and events. $5 donation to watch the carnage at the door, and spots still available to battle your own bot!
I’d been added to their mailing list after I’d taken a class on using woodworking hand tools for fine woodworking back in 2014.
Woodworking seemed like it would be a hobby I would enjoy. I had been on the stage crew in high school and college, building platforms for sets, rigging lighting, painting various flat surfaces and so on. I’d also worked for a summer between high school and college as a carpenter (more of a carpenter’s helper really), picking up trash and mostly sweeping around a build site, so I had what might be called “experience” with carpentry.
Since college, I’d done rough carpentry whenever I had the chance, which mostly was not at all until we got a house of our own in 2011, an American Bungalow built in 1932. Fixing different parts of our house keeps me busy most Saturday mornings.
My interest in fine woodworking lasted until I found out how much it cost to buy wood, or at least the kind of wood that was needed for the projects in Fine Woodworking Magazine. Building a bench out of cherry wood is much more expensive than just buying a bench made of cherry. Also, if I started making furniture, where would we put it? The aforementioned American Bungalow is only 1,500 square feet (2,000 if you include the finished basement), we do not have spare benches or other wood furniture.
So, by 2016 I was back in the market for a new hobby, though my immediate interest in the MASSdestruction Robot Competition was that it seemed like something fun to do with the kids on a Saturday. We are always looking for activities that get us out of the house, keep their interest for at least 20 minutes, and can be presented to other parents as vaguely educational. This Robot Competition thing seemed to fit the bill.
The event was great. The folks over at Artisan’s Asylum were and are very friendly and super welcoming to kids. The 2016 event may have also been when the first Valkyrie style robot was test driven by Alexander Crease, though I can’t be sure.
After the event, I floated the idea to the kids of us building a robot together and competing in the next MASSdestruction event. The kids and wife were on board. We began sketching different concepts for different types of robots and weapons.
I started researching on the internet how to build a robot, and came across crazy people who had machine shops in their garages and basements, and were fastening robots out of old drill motors and remote control car transmitters. That seemed like a huge overkill; I just wanted a project to do with the kids. I found the parts list for building a robot on 360labs.net, and bought a robot kit off of Servocity, a couple of Roboclaw motor controllers, a battery, and a transmitter and receiver , and started to put it all together. For the robot’s weapon, I created a drum weapon out of a length of pipe studded with some nuts and bolts I got from the hardware store.
I quickly learned I was out of my depth. It turns out that the rudimentary understanding I had of electronics was woefully under-powered to handle actual electronics.
Also, working with metal is very different than working with wood. Metal has much tighter tolerances. Most of the things which one would do which work out with wood, like drilling holes using a cordless drill, don’t work great with metal. The reason why is obvious in retrospect; metal is much stiffer than wood, and doesn’t take kindly to “forcing” it to work. What seems to be a perpendicular hole turns out to be off by quite a few degrees, which can end up mattering when you are trying to screw two parts together.
I ended up buying a set of transfer punches and a drill press which were sufficient for my needs to build the drum weapon. With those tools and a lot of trial and error I created my first combat robot, which was named Firestorm by family consensus. Over the weeks that it took too build, the kids lost interest in the idea, and building the robot became more and more a thing I was doing which the rest of the family giving me feedback.
With the new robot in tow, we arrived at a MASSDestruction event in January 2017 ready to do battle!
Or at least, I thought I was ready. My robot failed the safety test (!) the first time around, and I wasn’t sure why.
The safety test is pretty simple; you put the robot in a sealed box with a transparent lid, and then you have to be able to demonstrate 1) that you can move the robot around, and 2) that if you turn off the power to your transmitter, the robot’s weapon and wheels will come to rest by themselves. The reason I was failing the safety test was that after I spun the weapon up, when I turned the transmitter off, instead of the weapon coming to a controlled stop, the weapon would slow down, and then robot would start behaving erratically.
After I failed the safety test, I took the lid off the robot and stared at the circuitry, poking things with a continuity tester, trying to figure out if there was a shorted connection or something which was leading to the problem. When I couldn’t find anything wrong, I tried to pass the safety test again. The second time we passed, and were ready for our first robot battle (in retrospect, I was just lucky to pass the safety test the second time, as I had the a design flaw with my circuitry. More on that later.)
That was how I met Charles, who may not remember me. I admit to Googling Charles after the competition. His robot 12 O’Clocker is something to behold. I’ve studied the pictures of it and have seen it in action and honestly have no earthly understanding of how he get’s it to work. He has all the technical skills I lack and the courage to pursue his interests full time which I admire.
I’ve embedded a recording of the whole fight below. You can hear my wife and children cheering our robot on in the background on the video.
As you can seen, Charles’s robot crushed my robot in the first few seconds. The cardboard cover I’d made to “protect” the sensitive electronics inside went flying. My robot was a twitchy malfunctioning mess by the end of it, parts of it lying everywhere all over the ring. I had to scrape it off the floor with a broom. The whole thing was great fun and good entertainment for most present.
I quite enjoyed the experience, both building the robot and fighting in the competition. The only part I didn’t like was disappointing my son, who was understandably frustrated that the robot didn’t do better. The silver lining though is I think I read somewhere that you are supposed to let your kids watch you fail in order to learn that failure is okay and you can recover from it. The idea is that it help them build “resilience”, in which case this is all just good parenting.
For the next version of Firestorm (Firestorm II) I made a bunch of changes.
Gone was the flimsy cardboard cover, replaced by a sheet of steel. Gone was the drum weapon, which I could never get to work properly. I couldn’t measure the center of the drum weapons axis correctly, which meant that when I spun it up, it would wobble badly, adding vibration to the whole robot, which would start to unscrew the weapon from the chassis. I must have gone through five of six pipe end caps trying to hit the exact center; the 2-inch diameter back of a pipe end cap begins to feel as big as a football field when you are trying to find it’s exact center. Despite my best attempts to measure the center with a ruler, I couldn’t find it. I learned later that there are a whole bunch of techniques specifically focused on measuring distances on metal surfaces. Also, the drive train was ill-conceived, looping over the exterior body of the robot instead of being in a protected track.
So for Firestorm II, I replaced the drum weapon with a lawn mower blade mounted on the top and angled as an undercutting weapon, on the assumption that I would end up competing against 12 O’Clocker again, and that attacking 12 O’Clocker from the side and sweeping under it might increase my chances of success.
I soldered, zip tied, and super glued down every connection, wire and nut, learning from last time the importance of hardening all joints and connections against large impacts. I removed two of the original four motors to save some weight and allow for the new weaponry, figuring that two motors was good enough. I had aspirations to create multiple different weapons systems which I could swap out based on the different competitors which is why the lawn mower blade is mounted on an independent harness which can be removed from the chassis. Below I’ve embedded a video of Firestorm II in action for your amusement.
In 2018, I entered the next competition. Learning from last time, I did my own safety check the day before. My self-safety check failed, and I still didn’t know why. I spent that evening poring over information on the internet to try to figure out what was going on.
I’m sure any electrical engineer reading this is rolling their eyes at this point, but that was how I learned what a diode is and what it does. What was happening was that the momentum of the weapon was turning the weapon motor into an electric generator, feeding electricity back into the system. A diode prevents this by limiting the flow of electricity to only go one way. The advice to use a diode to prevent exactly what was happening was right in the RoboClaw manual (page 4. under the heading Safety Wiring) but at the time I first wired up the robot, I didn’t understand what I was reading, so when it worked fine initially, I thought everything was fine.
The morning of the competition I went to “You-do-it” Electronics Center out in Needham, bought some diodes, and in the parking lot patched one into the robot and dry-tested it without the blade attached. It seemed to work fine.
This time I passed the safety test. The folks at the Artisan’s Asylum scrambled to find another twelve pound robot for me to fight. They kindly hot glued something together, a two wheeled contraption which a rubber house and a knife attached which they kindly let me beat.
I don’t have footage this time of the battle, which all for the best really. My robot started driving in circles, and I was having trouble compensating by steering. The announcer thought it was the Coriolis effect of my lawnmower weapon blade spinning, but that wasn’t it.
I figured out later that I’d put the diode in the wrong location in the circuit. While the weapon was safe, at high speeds and with a lot of changes in direction, the wheel motors had the same kind of feedback issues as the weapon motor, which lead to difficulties steering. I eventually did manage to get in one good hit though, and the other robot fell apart.
It was trying to make Firestorm III which started me the rabbit hole of pursuing robotics as a hobby. I wanted to try to make a drum weapon again, and could not find one to buy or anything at the hardware store to adapt into a good one. A thread on Reddit educated me that you need to either machine one yourself of have one machined for you.
I dutifully downloaded from emachineshop.com their CAD software and designed a drum weapon in CAD. When I got back the quote from emachineship.com to have them machine one for me, it was about $360 just for the two parts I wanted. I said to myself “I could buy a mini-lathe and do it myself for that much”.
Ha ha ha how wrong I was, both for how much money it would eventually end up costing me to try to do it myself and for whether I had the skills I needed. Still, that is how late 2018 I ended up buying a Harbor Freight 7" x 10" mini-lathe, and went from thinking people who had a machine shop in their garage were crazy to becoming one of those crazy people.
