3D Model

Steampunk Owl Earrings

I was still very much infected with the steampunk bug at the end of 2016 when designing this for my girlfriend. Owls are cool, so I followed that animal route and poked around on google images in search of a design epiphany. I had a hard time envisioning the final model at the start, so I created each component in a piecemeal approach (over several days) to build up the whole ensemble. For the head, I played with Blender's sculpting feature and then tacked on a camera-like shutter straight from Aperature Laboratories for one eye and a radar dish for an ear. (If the picture doesn't load, I promise it doesn't look like a poorly-taxidermed owl out of Frankenstein's lab). I then added an armor/plate-mail-like surface on the body, and kept the wings simple to help reduce the weight. I didn’t quite finish it in time to be printed before Xmas (as there is a 2-3 week lead time for the printing/shipping), but this will give us a chance to calculate the weight for different material options and avoid a final product that’s too weighty. I want to end up with earrings that fall toward the middle of the weight distribution (as opposed of a couple standard deviations out on the heavy tail like the last pair).

3D Model

Brain and Head Layers

One of the components of my graduate research is to explore neuroimaging data. In our research, we need to keep track of the different head layers as part of a specific technique we use called “source imaging.” Source imaging is simply a mathematical tool that allows us to estimate which areas on the surface of the brain are active from data that was recorded outside the head (with EEG, for example). To use source imaging, we often collect MRI scans of people's heads in order to have a high resolution model of their different head layers (i.e., scalp, skull, and brain). With this model, we can calculate how electrical brain activity propagates through the different layers to our sensors. One day at lab, I stumbled across a one-liner command that allowed me to convert those MRI models into .stl files that could be easily loaded in 3D modeling software (like Blender). In other words, I could trivially load an MRI scan (containing extremely high resolution 3D model of my head and brain) into my 3D modeling software. That was too cool to pass up, so I loaded all the tissue layers and created a progressive cutout to illustrate how the head layers are modeled in source imaging.

Update: This image was part of Figure 2 in my 2016 paper and was selected for the cover of the Journal of Neural Engineering's October 2016 issue.

3D Model

Fish Pendant

This design was a birthday present for my sister. She wanted something to reflect the fact that she's a Pisces, so I tried to put something together to reflect that fact and also draw from a yin-yang symbol. The fish outlines were simple to do -- just some swooping Bezier curves that I filled and then used the Boolean operator to join. The fins and tails were very fun to design. I experimented with Blender's wave modifier and used a knife project to add the little notches at the back of each fin. Just like all the other models, I printed this one out of steel from Shapeways to give it some rigidity. The fish are purposefully wispy and wiry to keep the pendant from becoming too heavy. Metal is awesome to print out of, but (as I learned from my previous two projects) solid extrusions can quickly turn a design into a paper weight.

3D Model

Mechanical Elephant Pendant

Up to to now, this pendant, which was a 2015 Xmas present, was my favorite project to design. Through a long set of covert and disguised questions, I was able to deduce that my girlfriend would wear a steampunk elephant. There are some some really incredible mechanical elephant sketches out on the interwebs, so I poked around to get some inspiration. After making a few of my own sketches with different eye pieces, machinelike attachments, and gears, I stitched together the different components of the model in Blender and made a few renderings like the one opposite this text. I had the pendant printed out of matte bronze steel from Shapeways. (It still boggles my mind that it's possible to 3D print metal.) The matte bronze steel exudes a steampunk vibe, and it came out heavy enough to stand in as an emergency paper weight. It was strange to holding the printed piece though -- I basically got a strong sense of déjà vu when it first came in the mail. I spent quite a bit of time tinkering with individual vertices on my 3D model (in the virtual world), so holding the exact same physical object (in the real world) was weird. I imagine it would feel like "exploring" a city extensively in Google StreetView and then visiting in person a week later.

3D Model

Mārtiņa Zīme Earrings

This was my first real foray into 3D modeled art, and I really jumped into the deep end of the pool by using Blender 3D modeling pipeline. I had used AutoCAD before for engineering projects, but the desire to support open-source software (and lack of money for an AutoCAD license) led me to give Blender a shot. After doing 15 or so tutorials on youtube, I was able to design this model as a 2014 Xmas present for my girlfriend. She's Latvian and the mārtiņa zīme is a Latvian symbol with a cultural significance. Pinning down the meaning of any Latvian sign is a bit like nailing water to a wall, but it basically is a reminder to exhibit endurance through winter or other challenging times. The symbol is rotationally tessellated four times along the z-axis, and I eventually had it printed out of steel from Shapeways. In classic greenhorn fashion, I didn't consider weight when picking the material. They're apparently on the edge of wearable, but I need to reprint a scaled-down (i.e., lighter) version to preserve collagen in the wearer's ear lobes.


Infinity Table

This raspberry-pi controlled table produces what looks like multicolored tunnel that extends into the floor (despite the table top being only 3 inches thick). As I explain in the video, it relies on a strip of LEDs that sit between a normal mirror and a two-way mirror. Light progressively bounces back and forth (with a little escaping each time it's reflected off the two way mirror) creating the infinite tunnel effect. When I get a little free time, I'll will put together a more detailed construction tutorial and publish the python code on github for each of the 3 modes.

Custom-programmed modes

  1. Revolving rainbow
  2. Light organism battle-royale
  3. 2-team light organism battle-royale (aka, a glorified random coin flip)

Materials

  1. Raspberry pi
  2. 3V to 5V shifter and Raspberry pi hat to control LED strip. Code is modified from the tutorial here.
  3. 3 meters of 60 LED/m Dotstar LED strip from Adafruit
  4. 5V AC->DC power supply
  5. 80-20 T-slotted aluminum (check their surplus account on ebay for big discounts)
  6. Acrylic mirror (below LED strip)
  7. Two-way mirror for table top(sometimes called "mirror pane"). If you're in Seattle, check out Perkins Glass in Cap Hill -- I paid about 15% of what I was quoted elsewhere for the same piece of glass.


Video Demo


A Bit about Bitcoin

This is the Nerd Nite talk I gave, but the work-in-progress text version is below.

Presented by Mark Wronkiewicz @ Lucid Lounge and recorded by Nerd Nite Seattle.

Blurb: What is this bitcoin fad we keep hearing about? Hasn’t the news media, which always factually verifies its stories, reported its death like five times anyways? At this nerd nite, you’ll get a thoroughly geeky walk through of this resilient internet currency’s internal mechanisms, as well as the outlandish tales that have stemmed from its existence. Truth is truly stranger than fiction (especially when you’re talking about pseudo-anonymous internet money invented by some shadowy unknown internet figure that allows transfer of money between any two people in the world and is essentially impossible to regulate).

Mark is a graduate student studying Neuroscience at UW. His work focuses primarily on Brain-Computer Interfaces, i.e., controlling computers and robots by thinking for both rehabilitative and augmentative goals. Outside of the academic realm, he enjoys hiking, predicting stock price and sports outcomes, and playing the guitar (poorly).

One person's (inconsequential) Bitcoin experience

Near the close of 2014, I joined the herd of technophilic nerds on a modern-day gold rush. What we sought was not shiny -- in fact, our prize doesn't exist in physical form. And we didn't have to journey down a dangerous wagon-trail to get to the mining grounds -- all the traveling was done by information packets loyally obeying the orders dispatched by our computers. Instead of risking lives and livestock, the other starry-eyed miners and I risked rest and RAM; our tools were made of silicon, not steel, and GPUs, rather than horses, chomped at the bits.

The 2014-2015 Bitcoin rush lacked most of the physical components of its 1849 gold rush counterpart, but shared an eerie parallel with the psychological mania that swept up a group of susceptible young people hungry to strike it rich; people that know how to optimize their Mountain Dew intake over a 12-hour LAN party and had no problem properly seating a CPU on a new motherboard. Having recently started graduate school and undergone a baptism in Linux, computer construction, and all things open-source, I was swept along with the technophiles. The first time the Bitcoin concept really planted itself in my mind was while I was talking to Raeed, a good HS friend of mine, at the Society for Neuroscience meeting in San Diego -- a sort of mecca for brain researchers. It was there he told me that this monetary resource could be "mined" by diverting processing power from a personal computer when not in use. I was hooked right there while standing football-field-sized poster session floor at the thought of making easy money with a little electricity.

At that time, Bitcoin was blowing up on the news networks. Each journalist gave what felt like a compulsory explanation of Bitcoin's inner machinery in simplified and condensed terms -- an impossible task -- before getting to their true reason for writing: an explanation of how Bitcoin represented an economic come-to-Jesus moment or how the technology was a Ponzi scheme doomed to fail and ruin millions of people's financial future. I don't know, maybe they were getting warmed up for the hyperpolarized election cycle around the corner. Anyway, the writers with a technophile bent predicted that Bitcoin would save us from the next banking crisis and that its optimal design would obliterate coorporate greed in one fell swoop (with the swoop occuring over a 5-10 year adoption period). The other side, teeming with entrenched technophobes published reams of stories detailing bitcoin's role in propping up just about any nefarious economy known to the internet. Along with a few high-hatted economists, they couldn't finger-poke their keyboards fast enough in their efforts to detail its impeding implosion. Note that, so far, their accuracy is on par with the Mayan doomsday calendar.

Bitcoin's allure

The basics of bitcoin I'd heard before. It is a "currency" that exists in the digital world instead of the physical one. Some shadowy entity under they pseudonym "Satoshi Nakamoto" stitched together an impressive concoction of advanced computer science concepts that became Bitcoin. To the inner circle of the Bitcoin community, his name is synonymous with controversy. His/her/their story represents a favorite for those who have one foot stuck in conspiracy theory and the other planted wherever it is that leads to an uncontested nerd classification. It's unclear if Satoshi represents one genius computer scientist or a righteous group of programmers and there have been many interesting stories of people who were identified or came forward as the potential mastermind (see here, this, or this one). As of the time this was written (late 2016), the true owner of this glass shoe still isn't known.

The heart of Bitcoin's allure boils down to its decentralized nature -- no central manager or organization responsible for making the Bitcoin tick. To understand why being decentralized, let's look at the contrasting situation: a centralized system. Here, the obvious example is your run-of-the-mill bank. Banks store your personal credentials, manage all monetary transactions, and have some large, but probably countable, number of servers that tie the whole operation together. We call them centralized because a single company keeps track of all this information and we trust them to do so in a responsible way. But there are hiccups. Accounts of major companies are breached, executives setup shady internal policies (that will eventually get blamed on their underlings), and sometimes those at the top really, really need a new yacht and don't mind flushing the world economy down the toilet to get it (c.f., 2008 financial collapse).

With Bitcoin, no "trust" of any person or company is required (i.e., it's decentralized). Anyone can send and receive any amount of money from any other bitcoin account, and the network as a whole verifies every single transaction. In other words, there is no one company or person that is implicitly trusted. This is incredibly important because the "network" consists of millions of computers all with the same transaction ledger that are all cross-checking each other. If a bitcoin payment doesn't tick all the boxes, then the network doesn't recognize it as valid, and the transaction request goes into the digital ash heap. Oh, and the set of rules that all bitcoin users must follow? The code for bitcoin is all open source and always has been. Take a look at the code or the whitepaper. Any change in it's operation is completely transparent. This crowd-sourced and open-source nature nature makes it impossible (in any at all realistic scenario) to "hack" the network into making fraudulent transactions. Thus, you don't need to worry about a greedy CEO making poor policy decisions, or a countries government manipulating your currency. You also don't have to worry about natural disasters or an SQL injection that brings down a bank network. The information is backed up millions of times on every continent.

Of course, this is not all good. Because bitcoin is relatively anonymous compared to almost any other digital payment means, it's often used for less-than-legal internet business. Silk road, the eBay market of the illicit drug world is probably one of the most well-known examples. Many law-makers have expressed concern over the ease of using bitcoin for money laundering and other illegal purposes. Side note: the bitcoin technophiles would like you to know that traditional fiat money (i.e., cash) still serves as the backbone of the world's collective black market. Another downside is that the system Satoshi set up really puts financial responsibility into the hands of each individual. If someone hacks into your personal computer, gets your digital key, and routes your bitcoin to their account, you're SOL. If you accidentally send your money to the wrong bitcoin wallet when paying for your new bed sheets, too bad. If your house floods and destroys your PC containing the files for your bitcoin wallet (and associated digital private key), sorry. There is not and never will be a customer service rep to take your call.

What makes the network tick?

The bitcoin lexicon follows a traditional mining operation in the mid-1800s. Back then, you had "miners" who spent a lot of money up front to purchase all of their equipment (picks, axes, panning bins, etc.) and then continued to spend more resources (time and energy) to frantically dig in the dirt looking for some shiny stuff. A few pick their digging or panning spots in a smart way, but many were better described as frenzied masses who randomly search for pay dirt believing (almost religiously) that they would get lucky.

This terminology also underlies the bitcoin network. At a high level, the modern "miner" also spends a pretty penny to setup their equipment in the form of high-caliber computers. Once constructed, miners must spend energy -- electricity instead of elbow grease -- to make their computer go on this random walk to find bitcoins in the digital landscape. Every once in a long while, a mining computer happens upon the processes that lead to a sum of bitcoin, but this is incredibly rare. To mitigate the risk, modern miners usually band together in "mining pools" and distribute rewards when anyone in the pool gets lucky. Satoshi was smart though -- he harnessed this random walk to serve a higher purpose in the bitcoin network.

These miners really make the network run. That master transaction ledger we talked about earlier (containing a record of all bitcoins ever sent or received) is actually called the "blockchain." As people are sending bitcoins to one another on the network, these transactions build up for about 10 minutes in a short, temporary list. Miners are racing to solve the block, which means to add this list of transactions to the official block chain thereby cementing them in history (though there are sometimes weird corner cases we won't get into right now where this can be undone). To solve a block, a miner must take 1) a little text information from last solved block, 2) the list of pending transactions (also text), and 3) a small random bit of text, and run this input through a particular function called a "hash function." If a special result comes out, they have "solved" the block. Specifically, the resulting hash must have many leading zeroes, and this happens incredibly rarely. Solving a block comes with a prize of 12.5 bitcoins (currently)plus bonus transaction fees to keep miners motivated. It is in search of this prize that makes them tolerate the substantial electricity addiction of their mining computers that I came to appreciate in a very roundabout way.

Heating my home with mining computers

I finished my triplet of bitcoin mining computers in flurry of nonstop weeknight work sessions. Eight GPUs, three low-end mobos, a few flash drives to act as hard drives, and a batch of light-up, clearance case fans left me with three mining computers and enough cardboard boxes to fill a recycling truck. At the time, I didn't really notice that the weather was decidedly transitioning into the classic Seattle cold and rainy season. The doorways of my 1914 apartment also do not have the best seal against the outdoors and there is no flue in the chimney to prevent a backwash of cold air. Of course, the high draft-susceptibility index of my apartment was not apparent until I came home one day in early December and my living room was too cold to take off my winter coat -- it was probably 30 degrees cooler than the day before despite no shift in the weather from the day before.

Turns out the internet had temporarily shut off for some reason. (It's Comcast though, so I really shouldn't have been that surprised.) The loss of internet then crashed the mining software processes to exit because they couldn't send the output of their work to the mining pools. Therefore, the GPUs had stopped devouring the kWhs and quit serving as my stand-in, triple threat, octo-GPU space heater. A quick power-cycle of the router fixed the internet and put the mining operation (as well as my electric heater) back into top gear. It was also quite satisfying to know that I was stickin' it to the man; swapping out a computationally dumb space heater with eight GPUs that strained my apartment's aging circuitry but put this energy toward a smart and profitable purpose.

What does it all mean?

After the frenzy and late night work sessions began to die down, I was able to separate myself enough to mull over some of the philosophical issues at the foggy fringe of the Bitcoin network. For instance, where do these digital currencies like bitcoin stand relative to actual money? Despite being able to buy many actual goods from real vendors with bitcoin, why does it feel so susceptible to a crash. What exactly is "real" money anyways and how does anyone really know (beyond a few old fogies giving out ratings while dreaming of their surf-and-turf yacht party next weekend) which specific form of currency is more robust or reliable? On what grounds do said fogies or anyone else make a judgment calls on the value of a currency and how often must they recalibrate their criteria? What's keeping the U.S. dollar from following so many other monetary concepts into extinction?

There are so many questions, but I'm going to focus on one for now: Why is the confidence in the U.S. dollar a given for so many. A typical answer contained some flavor of the statement that U.S. cash is backed up with gold. On the surface, this made sense to me. The ability to trade out dollar bills for something recognized as valuable around the world is a comforting concept should the U.S. dollar collapse. Later I realized the fallacy of basing my knowledge of the gold standard on "Goldfinger" -- perhaps the best Bond movies ever made. In it Sean Connery is pitted against a ruthless European whose scheme was to build up a personal stockpile of gold right before irradiating the gold stores at Fort Knox with a stolen nuclear weapon. With the supply drastically reduced, governments would have had to bend to his every whim if Bond had not saved the day. Getting back on topic though, I found out after a little research that this gold-standard concept is as dated as that 1960s movie.

Gold-backed U.S. fiat ended with the Nixon Shock in the early 70s. The U.S. dollar is out on a limb so to speak with a more ephemeral backing made mostly of history and global confidence. But so what if it hadn't? Why makes the "gold-standard" such an attractive idea -- why do we even value AU more than almost any other element? Sure, it has practical applications in a few electronics fields and played a key role in the race to split the atom, but those are small beans. I could be wrong, but it seems like the real reason boils down to the fact that gold is shiny and rare. Long ago, someone equated "shiny" with "value," and the human race has been saddled with that belief for millennia. I hope someone will map the raccoon genome soon and calculate the overlap with human DNA.


iPhone App page is in progress!

Completed iPhone Applications

  • Braincopter: avoid oncoming planes and blimps by controlling your helicopter with your brain activity
  • DrugDB: quickly look up drug information. Aimed at doctors, med students, and patients.
  • PillDB: identify an unknown pill based on color, shape, size, and many other features.