One man’s fascination with visualizing big data could lead to better city planning.
Eric Fischer is a self-described geek, but that only scratches the surface of his nerdiness. The data artist and software developer’s geekery runs deep, encompassing data visualization, transportation, pedestrianism, tech history and maps — beautiful digital maps.
Fischer’s work has been featured on online news sites and displayed at New York City’s Museum of Modern Art. Last summer, Wired highlighted maps Fischer had created by tracking billions of tweets — and which mobile device was used to create them, by neighborhood worldwide.
Previously an artist-in-residence at the Exploratorium in San Francisco and a member of Google’s Android team, Fischer currently works for Mapbox, a company that provides users the tools to design their own custom maps. Fischer recently paused from his projects to discuss walking stats, massive data points and his favorite map of all:
How do you respond when people ask what you do?
I put dots on maps and try to understand what they mean, and write software to help other people do it, too.
What is Mapbox and how did you come to work for them?
Mapbox makes tools and services to help people design and publish maps. The main service is styled web maps, so you can have a map on your website that looks like it belongs there with your fonts and colors, instead of looking like it was transplanted from somewhere else. Early last year, as I was finishing up my video projection projects for the Exploratorium, Mapbox approached me to do some work under contract for the State of the Map US conference about OpenStreetMap. It worked out well.
Mapping race and ethnicity in Philadelphia, inspired by Bill Rankin’s map of Chicago, with data from the 2010 Census.
What is OpenStreetMap?
People often refer to OpenStreetMap as “the Wikipedia of maps.” It is a worldwide set of map data, along the same lines as the commercial map data from TomTom, Nokia or Google, but anyone can edit it to fix errors, and anyone can download the data to use for whatever they want.
What prompted the iPhone and Android maps that Wired featured?
They were part of a joint project between Mapbox and Gnip to show off the breadth of Gnip’s social media archives and Mapbox’s tools for mapping data. In addition to the phone brands, we also made new versions of the world languages and locals and tourists maps that I had made before, but on a worldwide scale and from a much larger set of data.
Which projects have you worked on since then?
I’ve been compiling and researching pedestrian count data to try to understand how much walking takes place, where and why, so that we can give satisfying walking directions and, in the long run, learn what is necessary to make better cities.
What kind of data are you looking at for that, and where does it come from? The data comes from the “turning movement counts” that transportation agencies and consultants periodically do. They are generally mostly interested in vehicles, but often count pedestrians. Most of these counts end up in reports about individual intersections, so I have been collecting as many of those reports as I can find and pulling the numbers back out to get an idea of what walking looks like statistically.
Mapping race and ethnicity in New York City, inspired by Bill Rankin’s map of Chicago, with data from the 2000 Census.
Is there info missing from walking directions that you’re hoping will be added?
The big potential that I see is in route choice at the street level, by understanding the trade-offs people are willing to make when they choose a more pleasant, more useful or safer route over a short one. We know from lots of past research that people like stores, trees, closely spaced entrances, narrow roadways and so on — but not how far out of their way people are willing, or even eager, to go for each of these things.
What’s the connection between walkability and quality of city life?
It’s hard for me to even think of the two concepts as separate things, although I know that’s a huge blind spot. At the same time, looking at the statistical patterns has also driven home that it’s never going to be possible for every street to have enough pedestrian activity to make it feel alive, because there just aren’t enough people with enough time to fill all the streets.
Your Twitter bio says something about misspelled street signs … It’s really two separate things, but you only get so much space on Twitter: misspelled street names in the sidewalk and old street signs in general. I think my interest in the street inscriptions dates to wondering in childhood about the leftover “Bellefontaine Street” written in the sidewalk near where I grew up, an artifact of a street renaming decades earlier. And then I was encouraged to keep looking for them in San Francisco after finding the Cement Mix Up group on Flickr.
Is there something that drives you nuts if you see it on a map?
Everyone has their own pet peeves, but my complaint will always be about colors that I can’t tell apart. I’m red-green colorblind, and I think part of my distinctive style might be that anything that has enough color differentiation to be really clear to me will look a little wild to people with normal color vision.
But the prominent colors in your iPhone and Android map are red and green.
The particular hue and brightness of green in that map actually works reasonably well for my eyes, even though it looks like it shouldn’t.
Mapping what has been seen (via Flickr) and said (via Twitter) around Boston. (Red dots are locations of Flickr pictures; blue dots are locations of Twitter tweets; white dots are locations that have been posted to both.)
What advice would you give to someone who’s working with interesting data and wants to make it visually appealing?
Paying attention to dot size, density, gamma and color mixing and taking advantage of the full range of brightness and saturation will get you a long way.
How do you deal with billions of data points?
In many ways, it’s actually easier to work with large data sets than small ones because with small data you’ve got to squeeze everything you can out of the tiny bit that you know, while if you have a lot to work with, the patterns emerge more naturally. The downside to large data is that you have to be much more aware of the memory and time that it takes to process it all.
Do you have a favorite map that continues to amaze you every time you look at it?
I’ll nominate the 1857 U.S. Coast Survey map of San Francisco for its wealth of historical detail, and as a monument to the precision that can be achieved without any digital technology at all.
Images courtesy of Eric Fischer.
A professional writer and editor, Alyssa Danigelis focuses on the intersection of technology with sustainability, business, media, arts, and design. Her interest in technology has led her to cover self-healing power grids, 3D-printed food, wearable computers, and robotic couture. Originally from Vermont, she’s a graduate of Mount Holyoke College and Columbia University’s School of Journalism. She lived in New York City for several years before falling in love with sunny Boulder, Colorado, where she currently resides. Her writing has appeared in publications that include MIT’s Technology Review, Natural Health, Fast Company, Inc. Magazine and Discovery News. Find her on Twitter at @adanigelis. She’s excited about sharing her passion for cool tech with iQ by Intel’s readers.