MIT’s Charles Guan on Constructing the LOLrioKart and Electric Vehicle Technology
Regular readers of “The Digital Student Blog” know first hand our fondness for technology. That is why we were so smitten with Charles Guan’s LOLrioKart, the electric powered shopping cart that has been seen tooling over the asphalt in Cambridge.
After learning about the cart at Gizmodo, we had to call our readers’ attention to Guan’s whimsical attempt to strap a set of large nickel cadmium batteries and a massive brushless motor to a derelict shopping cart. But our own fascination with technology and moving vehicles had us yearning for more – we simply had to hear firsthand from the creator himself.
So we spent some time with Charles, he of the web site Equals Zero with URL etotheipiplusone.net (yes, that’s Euler’s Identity buried in his call letters), discussing his recent fame, the MIT campus culture and his desire to construct something, anything when given such wonderful parts. We think you will enjoy some of the lessons Charles has learned from constructing the LOLrioKart including dealing with steering slop, his Fred Flinstone braking system and the initial construction of a vehicle featuring only two speeds, zero and fast.
The folks at Gizmodo gave you some great publicity but they also called you a nerd on their site (seeing a motorized shopping cart with a nerd inside tooling around on the street). Are you OK with that, I mean their site is read by somewhere near a gazillion readers?
Pop culture comes up with new ways to describe “someone smarter or craftier than you” every once in a while, but the classics seem to be ‘nerd’ and ‘geek’. Besides the occasional (and inevitable) grade school hijinks, I don’t recall ever really being victimized or shunned when referred to in that fashion. Part of getting over that is acknowledging the reality of the game – that what you do or who you are is not what the majority of people are used to. It’s useless to be sensitive about every descriptive term that comes your way, especially since they evolve so quickly, and people are multifaceted enough that anyone could come up with a word and pretend it’s derogatory or obscene. I mean, just take a look at the sheer number words that we have turned into sexual slang.
Speaking of which, just make sure you’re careful to not call someone a dork. In reality, you are indicating that they are a whale penis. Whales tend to be larger than you in every single way. I’ve been unwittingly complimented in this fashion many times.
Just so we know, what in your eyes is the difference between a geek and a nerd – does it have anything to do with the desire to strap technology to something that moves?
I have never really discerned between ‘geek’ and ‘nerd’. The running joke between my friends years ago was that a geek was someone skilled or passionate in a specific, tangible field of interest or hobby – computer geek, band geek, robot geek, etc. the list goes on. We sort of chose to designate “nerds” as being the underclass – those who are incredibly, almost frighteningly smart, but in a purely academic sense.
This definition is clearly not the universal one, some may argue it’s the exact reverse, but you know what? I don’t really care. Labeling me as one thing or another doesn’t really impact how I do things, and it shouldn’t affect you either.
Can you talk a bit about your first “attempt to strap a massive brushless motor to something?” I believe you called the electric vehicle: Snuffles the First?
Back in the summer of 2007, I came back from a trip to China with what I considered at the time to be “exotic equipment” – that is, brushless motors, precision-made actuators, lithium batteries, etc. The stuff that R/C hobbyists used to drop big money on but is now incredibly affordable, except my version of “affordable” was an order of magnitude or two lower. Keep in mind that for years before this I’d been hacking together robots with cordless drill, remote control toy, appliance, and other random parts, so this was pretty damned exciting.
One of these “exotic” parts was a large model airplane electric motor. Better-made equivalents from the depths of Germany and Switzerland are used to fly electric models with wingspans of 20 or more feet. Absolutely enormous. I didn’t really have anything that large (or really anything) to fly, so I sort of mulled over it wondering what I should do.
I frequented an area flea market, naturally to search for prospective robot parts. One of the usual toy vendors that day had a small electric scooter – the Sharper Image ElectricX2, roughly Razor scooter sized. Needless to say, when I saw this thing, the proverbial light bulb probably went off with enough intensity to start fusing hydrogen. It was totally trashed, the batteries were gone (and the guy had no charger, manual, or other support equipment), so I got it for a few bucks.
The first thing I did when I got home was tear everything down for inspection. When I discovered that the airplane motor could swap into the stock scooter motor’s place with minimal modification, it was pretty much on.
I kept all the build reports for that scooter on my website if you want to browse through it.
How successful would you categorize that endeavor?
While the execution was shaky – literally, because I didn’t have access to machine tools like I do now, it was absolutely a success in how much it taught me about Electric Vehicle technology. Stuff like what batteries would work the best, what motor controllers, and places to mount components. Spec-ing out these things for an EV is a lot different than for a combat robot. It was what threw me into the world of EV and hybrid technology that I’m sort of pursuing now.
Since the scooter was actually wired up and finished a few days before I moved to Massachusetts, I brought it along with me to campus. It was hilarious, and a big hit with the other frosh. People never really expected an electric scooter to kick them off like that.
The other way it was a success was that it landed me my first (and still ongoing) on-campus job as an undergraduate research assistant for Smart Cities. They posted an opening for a team that will be designing and building a light electric vehicle, a (street-legal, Vespa-style) electric scooter. Vespa…Razor… close enough, right?
Part of what I enjoy about engineering is the degree to which you can engage yourself in learning a new skill. You HAVE to DO it. It’s not something that can be read out of a book and repeated. If you measure success by how much you have improved your condition, then Snuffles the First was one hell of an accomplishment in that it taught me things which I can’t get from just sitting in class.
From your site I gathered the Nickel Cadmium batteries powering the LOLrioKart are not your ordinary flashlight batteries. How powerful are these?
The large-format NiCd cells I ended up using in LOLrioKart were donations to our club from the Solar Electric Vehicle Team years ago. They were pretty tired batteries. Each pack comprises 11 cells, for a nameplate voltage of 13.2 volts. Problem is that nickel batteries are prone to developing internal crystalline parasitic structures that cause quick self-discharge, so many of the cells were just plain dead. I used a procedure called “zapping” to revive them. It pretty much boils down to injecting a massive high voltage and current spike into the battery in an instant, such that the crystals just vaporize.
Kind of like a heart defibrillator, except the “vaporize” part.
When new, big NiCds like this can easily dump a thousand amps in an instant and maybe a hundred or two amps for a while. For the uninitiated, 1,000 amps at 54 volts (the voltage of LOLrioKart’s 4 packs when charged), is 54,000 watts, which is very roughly the same power consumption as a thousand 60 watt household light bulbs, or about 72 electrical HP, if it were actually doing something useful.
Note that “instantaneous” is the key here. It takes only about 500 watts to keep the kart cruising at a constant speed. It only takes about 6-10 horsepower to keep a *real car* cruising at a constant highway speed, on flat ground.
Realistically, I think the cells have deteriorated to the point of only being able to burst-discharge a few hundred amps. Still not something you want to drop a wrench across. I really don’t have proper facilities to take care of batteries this large (chargers and battery managing equipment get very expensive at this scale), so all I’ve been able to do is charge and discharge them slowly – which doesn’t really help with reviving them all the way, since fast charge and discharge cycling is what helps batteries stop being stale. Regardless, the rest of my electrical system will not handle the cells at full flow.
So, if Ahmadinejad had access to these Iran probably still could not go nuclear?
Ahmadinejad could probably use one or two to power his lights in case he has to go hide somewhere, but he’ll need a few hundred thousand more if he actually wants to kick a reactor into shape with them.
What was the impetus for the name, LOLrioKart? Where there any other names that you seriously considered?
It was incredibly spontaneous. I had no idea how the project even spawned, but the name ‘LOLrioKart’ was the first and only name, inspired of course by the MarioKart games. Adding the scrambled syllable takes after the classic internet meme of “lollifying” things. You may have heard of “LOLlerskates”, “ROFLcopter”, “LMAOplane”, et al. from a few years back. That, too, was sort of
How much practice did it take to be able to drive that puppy? In the video you make it look easy. Isn’t the center of gravity a tad high with you inside?
When you’re the builder and overseer of something, you kind of naturally get used to it due to being constantly with the construction and testing of the project. Piloting the kart during test runs wasn’t all that difficult when it was under control, but there were definitely a few moments where I thought somebody else was going to have to call home.
The kart’s center of gravity is actually very close to the ground. The battery pack in total weighs about 120 pounds, and it is all in a solid mass that is 3 inches above the ground. The motor and transmission in the back add another 30 pounds that isn’t much higher. I am not a heavy person, so sitting on top of all that probably raises the CoG only a few inches above the wheels. The kart does have substantial tilt when cornering at high speed due to the tires compressing, which amplifies the “I am about to roll over and die” effect.
How about the steering mechanism?
The kart steering is a simple “drag link” system commonly used on go-karts, except more aluminum-y. There’s no rack and pinion or hydraulics. It’s a simple mechanical linkage that transmits steering wheel movement one-for-one to the wheels. This means it’s very squirrely on the handling, and I had a pretty bad problem with slop in the whole steering shaft on an earlier build of the front end. This slop would cause the front wheels to not match the steering wheel angle often up to +/- 5 degrees. That doesn’t sound like much, but now just driving your car pretty fast and suddenly twitching the steering wheel a quarter turn. Then a quarter turn back. Not something you’d want on a busy street.
How do you actually stop the kart?
The short answer is that I stop it by willpower alone. Originally, I designed in a set of mechanical, cable-actuated band brakes for the front wheels. Electric motors have this cool feature that you can use them as electric brakes, and at the same time capture power back into your batteries. This is known as “regenerative braking”, and all the cool electric trains and hybrid cars do it. Because of this, I already had a really big brake on the rear axle – the drive motor. Adding a regular “sprocket brake” on it, like most simple karts have, was just redundant.
Problem was that these band brakes I spec’d out were for children’s toy electric scooters. The kind that go about 5 miles per hour. Long story short, they sort of vaporized on the first stop. I now know that you can get disc brake packages for mountain bikes and larger electric scooters. Guess what is coming in the mail soon?
So firing this baby up was truly a walk on the wild side?
The culmination of all of these safety non-features is that at one point in time I had a brakeless kart with a completely loose and sloppy steering linkage that was capable of exactly two speeds – zero and fast. I had recently blown a custom motor controller that I had built, and out of frustration, put the kart on “contactor control”. That’s just a very fancy way of saying “touch the battery wires to the motor.” On all 54 volts, the thing would kick up and wheelie on start and hold the front end in the air for two seconds, reaching some absurd speed in the meantime. After which, of course, I couldn’t really *STOP* the thing, and it would occasionally take off in a only slightly tangential path due to the steering slop. To slow that version, I would either just Flintstone it or perform an analogous maneuver that airliners use when approaching a runway and need to burn off speed – make little S-turns.
But no major crashes to this point?
The closest that I ever came to a real crash was when I made the mistake of closing the circuit making a turn. When the front end of the kart is off the ground, the steering is mysteriously missing. As soon as it crashes back down, of course, I am still attempting to be in the apex of a turn. The combination is that I sort of take a path midway between the two, which conveniently led to a curbside. It was a very elastic collision and I was certain the kart was going to catapult me out onto the ground, but somehow I landed back in.
That impact took out the right wheel and bent the 1/2″ hardened steel bolt that was the axle stub for it. The upward force of jumping the curb pretty much trashed the steering linkage and steering wheel mount. That’s when I finally had a good reason to rebuild it – the new setup includes a right-angle transmission that one, got rid of the slop and two, allowed me to add a gear ratio to the steering. It wasn’t much – only half a turn of the wheel from lock to lock, but oh boy was it a godsend compared to a quarter turn plus or minus an eighth.
I am still alive. Not sure how, but it feels okay. The two videos that I shot were well after these incidents, after the (variable speed) controller was refinished and the steering linkage made more solid.
But don’t mention the brakes.
Don’t know if you saw the results of the British study regarding the importance of what a man drives when it comes to impressing the ladies? What kind of reaction do you get from folks when you buzz by them in the rio? Is it attracting looks from members of the opposite sex?
One of the more depressing misconceptions about MIT is that, like many tech universities, it has an unbalanced gender ratio. This is patently untrue, and I am personally glad, since the actual ratio is near parity. The kart gets its fair share of attention from engineering women, but then again, I haven’t really tried using it as a chick magnet. I would attribute more of those reactions to novelty and “what on earth…” than anything else, though. Pretty much all of my female peers know about its existence and construction. A majority of them actually want to try driving it themselves, but due to the, ahem, health concerns, I have yet to let anyone else try.
General reactions to the kart have been diverse. If you’ve seen the videos, you have clearly seen people who don’t seem to notice it. Quite a few comments on other places where the kart has been featured seem to say that MIT students are jaded or ‘used to’ weird things flying/driving/crusing around. I would say that this is at least partially the truth, since there are many projects here which move around, and therefore the kart, while funny, isn’t really that weird.
For comparison, the DARPA Challenge guys occasionally take their autonomous forklift for a test spin. This is a 10-foot tall full-size pallet lift with dozens of sensor boxes, computers, spinning LIDAR rangefinders, cameras, and green underbody lighting hanging off it. Let me emphasize the fact that it is AUTONOMOUS. You don’t want to piss it off.
How about if you take it off-campus?
The reactions on the streets of Cambridge are much more varied. This is unsurprising, really, since outside of Harvard and MIT, Cambridge is just a suburb of Boston. I had more interested looks and people trying to catch the kart on their cell phone cameras when I wandered off the confines of the campus buildings.
The whole thing sort of reinforces my feeling that colleges in general tend to be little idyllic bubbles. The problem comes when people try to extrapolate the knowledge they gain in this bubble to the world at large, some times recklessly, in my opinion. I have gotten the impression that adults often think that college students are idealistic and naive, and I guess I can’t blame them. However, that
gets into politics of a completely different subject matter.
Any issues with the campus or Cambridge police? I wondered if someone might have been a bit unnerved at you bombing by them in this puppy and decided to call security?
At the end of the day, the campus police are around to maintain order and to make sure we don’t mire ourselves too deeply in the legal quicksand. Keeping people in line and enforcing every rule? Not so much. They’re pretty much all chill folk from my experience, and I’d conjecture that little oddities here and there make the workday all that more interesting. I have never yet had a problem with the CP with any of the club’s shenanigans.
I haven’t had enough encounters with Cambridge Police to really give an opinion, but have yet to get flagged or pulled over. My older peers tell me that they don’t mind unless you get really egregious (e.g. I start blowing red lights or driving the kart to work every day), because the history of being around MIT means It Can Happen Here™.
I tried to stick to every traffic rule that I could on the test runs. Not having turn signals, I used motorist hand signals. I did stop on red lights, and went on green ones. I would think at least part of the reason nobody ever called it in was because of this visage of sanity – instead of, you know, flying around at random.
So no speeding tickets? Hey can the cart really do 45 mph?
Nah, speeding was never a problem, since the kart maxed out at around 30 mph. Originally, yes, it was geared for the full 45, but I switched to a smaller sprocket on the motor because it… was, well, too suicidal.
I like my projects only *somewhat* suicidal.
What’s on the horizon for the LOLrioKart? Any other ideas about strapping brushless motors to other objects or other ideas percolating that we should be looking for down the road?
Up until now I considered LOLrioKart a terminal project – finished and done, time to move on. However, with the attention it has gotten, and the encouragement I’ve received from professors and peers, I’m going to see what part of my research I can integrate into it.
Remember that the work I do with the Media Lab is centered around electric vehicle technology. Part of what we are working on is a small, “foldable” compact urban vehicle. It features a lot of cutting-edge goodies, such as integrated in-wheel electric motor modules, steer-by-wire, lithium nanophosphate batteries, one-piece passenger shell, etc.
One of the ideas I had for the kart was constructing four wheel modules – each with their own internal motor – and mounting them to an updated subframe that has a suspension. The biggest shortcoming of the kart is that it’s a solid frame, which means every pothole they pave the streets here with goes straight into me. Not very fun.
While this would just be a go-kart with independent suspension, I also want to incorporate steer-by-wire into the wheel modules. The innovative feature in the car project is that each wheel can actually swivel up to 120 degrees, controlled by electronics. This is best explained with a picture.
Anyway, the bottom line is that I want to have a kart with four independently suspended, driven, and steered wheel modules. Fortunately, all the knowledge and experience I’ve been able to pick up because I work for them has gotten me to the point of being able to fabricate my own 3-phase brushless DC motors for the wheel modules, given some starting material (e.g. the hard-to-make-and-expensive-to-have-made laminated iron core that resides in most motors).
I’m pounding on the electronics side pretty hard now to learn how to make a functioning AC inverter for driving those motors. What this entails is essentially three of my custom DC motor drivers back-to-back, all controlled in a fancy cycle by one microprocessor such that I can do what amounts to moving something with my mind. Then I need to make four, then network them all.
Pretty freakin’ tall order, eh? Hell, I might even try to get this graded or something.