My final report for the DREU program was also submitted as a research paper to a conference. The report will be made publically available after the blind review is complete.
This is the first year that CHI has had two deadlines, one to produce meta data to start the writing process, and a final deadline next week for the whole submission. It is interesting how stressful 150 word abstracts can be if you are committing your months of research to those words. It will be interesting just how many papers that were submitted with metadata retract by next Wednesday.
My work is still primarily focused on writing for the CHI deadline. We have frozen development on the system we are describing. Each of the demos in the paper runs smoothly and I've been able to capture footage of them for a video figure we will submit along with the paper. As we get closer my focus is on getting the work into as many hands as possible, and having fresh eyes read and understand the contribution.
My time for the next few weeks is entirely dedicated to CHI 2017 paper writing. I'm primarily focused on my work on 3D modeling and modularity. The focus of this work has narrowed significantly and we have a full (but a little too long) draft ready for the deadline. This seems like a good place to be, as its always easier to cut text rather than to create it. I'm coauthoring two other works, and dedicating an increasing amount of time to those works.
We are now very close to the CHI deadline, within a month of the when the final papers are due. This has the department rushing around and I'm working very hard to build a succinct story around my research into software engineering and its applications to 3D model design. I've been meeting with researchers at the SEI to build a bridge between HCI technical writing and a software engineer's perspective.
In my last DREU experience I was able to coauthor a paper where we reviewed assistive technology on a popular 3D modeling website, Thingiverse. While that work was an instigating factor in all of my current research, it has cursed me to forever review thingiverse in my future research. I'm currently working on two projects that require me to examine the Thingiverse website in search of modular designs.
This week I've picked up more knitting skills. In order to understand the graphical structure our compiler must produce I've been learning to knit from a book of swatch patterns. Its an interesting way to keep getting research done while attending lectures or sitting in on meetings. I'm starting to build an intuition about how textures are developed and what needs to be tracked by our system to produce interesting graphical results.
I've primarily been describing the knitting project in my last few posts. But half of my time is spent working on software engineering research focusing on how to bring modularity to 3D design. This work has been ongoing research for months, and has less absolute breakthroughs on a weekly basis. However, the core focus of the work is coming in line. Which is good considering the ever approaching CHI 2017 deadline.
Because I've focused on research opportunities through out my undergraduate degree, I've taken an unusual route through my course work. Because of this I've missed a few quintessential course including compilers. For the knitting research however, we've decided that a compiler is just what we need: knitspeak->a graphical structure->(graphics, machine code, and improved knitspeak). Because of this I'm taking a crash course in compiler design, narrowing in on a particularly weird case: converting a natural language to a programming language.
I'm new to knitting. I've been a crocheter for many years but had never picked up a set of knitting needles until I started this project. This has given me a unique opportunity to have an outsiders view on the process. Rather than starting as a hobbyist, with specific projects in mind, I've been able to learn while studying the computational properties of textiles. This is an odd way to learn knitting, but I think it may actually have made me a better knitter faster. I quickly noticed that knitting is just a direct graph, loops connected to more loops, with a topological ordering that is decided by yarn. It has an almost analogous translation to computer data, its a 3D structure that is entirely built from a manipulation of 2D data.
I'm an undergraduate researcher from Colorado State University working with Dr. Jennifer Mankoff (Jen) at Carnegie Mellon University through the CRA-W Distributed Research Experience for Undergraduates. My research focuses on accessibility and rapid prototyping. Right now I'm working on two projects. One is a continuation of prior work where I'm developing a modular prototyping paradigm for CAD tools. The second, new project, is a knitting compiler that can convert natural language knitting patterns into computer code that can be edited or sent to a knitting machine.