For 24 hours during the 2014 MRS Fall Meeting in Boston, 14 students, professors, and postdocs took part in the world’s first MatHack Materials Hackathon. The event was hosted by MRS and sponsored by Citrine (www.citrineorg.wpengine.com). Below you can learn what a hackathon is, see details about the teams, and find links to the open-sourced code from the event.
But first, I want to editorialize (this is a blog, after all). When Bryce, my cofounder, originally pitched the idea to me to host the hackathon, I was spiritually on board, but thought he was a little crazy. Hackathons are for programmers, not materials scientists. MatHack demonstrated that hackathons are really about people. A hackathon is more about cultural experience than an unusually intense love of Linux. It is more about shared suffering and support and pursuit of a goal than it is about bugs and compliers and code. The MatHack teams came in with varying levels of coding talent and uncertain expectations, but a displayed a clear passion for solving materials problems with software. The work these teams produced was scientifically impressive, but more importantly: they came in with only ideas and finished with products. They worked together with people they had just met around shared ideas.
MatHack created a true scientific community. In an age when there is much hand wringing about how hard it is for young students to get into science, particularly those from underrepresented populations, events like MatHack can make science seem more exciting and practical. Credit goes to those that took the time to take part in the event. They are the ones building the materials science community from the inside, and I was lucky enough to be in the room to see it happen.
We are already planning the next MatHack and have incredible interest from all sides. I can’t wait to see what comes next. If we can maintain the seed that was planted at this year’s event, I have no doubt that this interactive community-building is going to lead to great things for everyone involved.
What is a hackathon?
A hackathon is a short, sprint programming competition where people come together to ideate, create teams, build a solution, and present their work, usually in 24 hours or a weekend.
Step 1: Ideate
Teams arrived in The Hub, the central area of the Fall Meeting, and gave 30 second pitches as responses to a very broad prompt: “Build a piece of software that would be useful to a materials scientist.” Some of these projects, as you can see below, probably deserve their own PhD projects, others tried to solve everyday problems that come up in working with materials, and yet others provided a new means of outreach to foster the next generation of diverse, engaged materials scientists.
Step 2: Form Teams
Each pitch was given a table and participants coalesced around ideas that inspired them. Most participants had never met before, and built relationships over the course of the 24 hours. The teams that formed were the real power of this event. While the software output was stunning, bringing smart, talented people together around materials is the reason the Materials Research Society exists.
Step 3: Build
Once teams formed, it was time to get to work. In some cases, people had an idea of what to do from the outset; in other cases, they had to learn a totally new programming language. Gallons of coffee and pounds of ice cream, nuts, and snacks fueled the teams through the night at the conference center.
Step 4: Present
At the end of a long night, participants presented their ideas to a panel of judges from academia, industry, and national labs who selected the top three teams for cash prizes. At presentation time, most of the faces were tired, but no one seemed like they were ready to stop working. Bruce Clemens, Stanford professor and former MRS president, probably said it best in the judging room when he noted, “These teams blew me away!”
First Place: Phone-on Flow
Members: Nicole Adelstein and Andre Schleife
Team Phone-on Flow built a functioning 3D crystal viewer. Using an Android phone, a student could visit a special webpage and, using Google Cardboard (which costs less than $20), could take an immersive virtual tour through a crystal structure. Even to the judging panel, who all have seen their share of crystals, the visualization was inspiring. As an outreach tool, such a low cost, interactive project could open up materials to a new generation of students. The team plans to continue to develop the site further and enable more and more robust features.
Second Place: MatHack QuasiCrystal
Member: Wenhao Sun
Wenhao Sun went solo for his project to develop a new way to use Density Functional Theory (DFT) to simulate quasicrystals, which can be thought of as the material equivalent of the number π: perfectly ordered but eschewing a repeating pattern. More importantly, they are a class of materials that shows a lot of promise for many applications but to date has not been thoroughly studied by DFT. MatHack QuasiCrystal is a first step in being able to simulate these materials to understand them better.
Third Place: Directed Materials Design
Members: Andrew Long and Ioan-Bogdan Magdau
Team DMD MatHack took an Operations Research approach to understanding materials. By looking at large-scale datasets, Andrew and Ioan built machine learning models to predict how various perovskite materials would work as water splitting materials quickly and without requiring expensive supercomputing time or the upfront investment of experiment.
Members: Sabrina Ball, Brendan Nagle, Katie Van Aken
Team CIVR started with a clear question: “How can I automate the solving of equivalent RC circuits for electrochemical systems?” For the experimentalist, this is a hard problem. It involves solving huge sets of equations and fitting them to experimentally measured I-V curves.
Member: Guoqianag Xu
MacHack is a tool to analyze the grain structure of a material using optical microscope images. It has a different take from most approaches: it uses multiple images of the same spot under different lighting conditions combined with machine learning to identify grain boundaries.
Members: Gabriela Correa and Oleg Rubel
Team MatMod wanted to address the problem that DFT (see above) is very computationally expensive. This is no small feat – Walter Kohn and John Pople shared in the Nobel Prize in 1998 for the development of the method and for operationalizing it in working software. MatMod’s approach was to use alternative k-point sampling techniques to reduce the complexity of DFT calculations for any particular system without substantially reducing accuracy.