David Hoffman's bicycle is still there, splayed in the street at the intersection of Penn and Negley where a silver sedan clipped him at 25 mph on his morning commute to Downtown. Hoffman's injuries -- to his face, wrist and chest -- have healed, thanks in part to $800 at the emergency room that January morning four years ago. His bike, too, has recovered. But it still lingers at that intersection, as do the words Hoffman's offender left him with, driving away without so much as stopping: "That's what sidewalks are for!"
It's an accident well documented in Pittsburgh cycling lore: It was the catalyst that brought Hoffman to quit his cushy computer job and form Bike Pittsburgh, a two-wheeled advocacy group. But what Hoffman didn't know at the time is that his accident was part of a Penn Avenue tradition -- hit-and-runs are to that area what "doorings" are to Squirrel Hill. It's the kind of colloquial knowledge cyclists share while backpedaling at stoplights or lounging in coffee shops -- or, since October, on Bike Pittsburgh's online interactive Bike Crash map at Maphub.org.
Click on "Bike Crash," and Maphub's default layout -- a detailed street map of Pittsburgh -- is sprinkled with small bicycle icons, each representing a real accident input, by address or GPS location, by anyone who has signed up for the free hub. Click again, and each icon reveals details about the accident: All the information about Hoffman's accident -- from the location to the hospital bill -- can be gleaned from the accident's icon. Click again, and registered users can share their comments on the incident, from a heartfelt "bummer" to advice on avoiding accidents in that spot.
The bike-accident map is just one of more than a dozen interactive "hubs" currently active on Maphub.org. Each hub offers a separate overlay upon a Pittsburgh street map, and each overlay contains icons -- and associated information --added by that hub's users. Other Bike Pittsburgh hubs, for example, display icons for the group's proposed system of bike lanes and completed network of bike racks, as well as members' tips for safely traversing the city's bridges. Click on the "Panopticon" Hub, and you'll see icons denoting surveillance cameras in the city; the Dumpster hub displays users' favorite Dumpster locations. With both hubs turned on, you'll get an idea of the best surveillance-free, Dumpster-diving options in town.
"[Maphub] is useful for us on two levels," says Hoffman, who still works with Bike Pittsburgh as director of member resources for its national counterpart, Thunderhead Alliance. "It's a fantastic resource for novice cyclists -- where things get tight, where it's dangerous, where amenities are." Clusters of icons near Penn Avenue or Squirrel Hill, for example, affirm the wisdom of taking side streets in those areas. But Hoffman adds that there's another use for the information: "[I]t's historically very difficult to capture bike-accident statistics -- as few as one in 10 bike accidents nationally are reported. So this could be a good starting point for historical data for the area."
"The goal is to provide a tool for bicyclists to communicate," says Bike Pittsburgh Executive Director Scott Bricker. In fact, some 200 users have signed up to be members of the bike-accident hub alone. But Bricker also hopes it will "effect positive change within the city" by showing "where it's obvious that we need new bicycle parking, or where we might want a bike lane."
This kind of thinking is music to the ears of Maphub's creators, Carnegie Mellon University graduate Nathan Martin and doctoral student Carl DiSalvo, and programmers Jeff Maki and Evan Merz. It's hardly the destination they could have charted for themselves when they first conceived of Maphub more than five years ago. But in recent years, particularly with the launch of tools such as Google Maps, online mapping has become big news. And Martin, DiSalvo and Co. have found their Maphub project changing from an art piece into a powerful tool -- one capable of hacking directly into Pittsburgh's communities.
Martin and DiSalvo are members of the art collective Carbon Defense League. Since its inception in 1998, the CDL has been best known for hacker-inspired "tactical media" projects -- projects which use technology and media in pranksterish ways, usually to communicate political ideas.
One project, for example, instructed punk-show attendees in the basics of rewiring the Nintendo Gameboy video-game system, creating a new game in which the player navigates a truant schoolkid's world, searching for a brothel. (It's not as seedy as it sounds: Super Kid Fighter, as the game was called, was based on the writings of psychoanalyst Wilhelm Reich regarding sexual rights.) Another CDL project was a Web site providing printable stickers with UPC-ready barcodes for low prices. If the superstore item you want is beyond your means, just print out "$9.99," stick it on the box, and pay the discount instead. More recently, CDL authored the Web site FtheVote.com -- a 2004-election Web site where lefties purported to offer sex to conservatives who promised to vote against Bush.
DiSalvo and Nathan Martin don't cut particularly punk-rock figures: DiSalvo is wearing horn-rimmed glasses and cardigan, while Martin is clad in hiking fleece and boots. It's much easier to imagine them doing what they were doing five years ago: working cubicle-by-cubicle at the same San Francisco design firm.
Before that, however, DiSalvo was living in a co-op in Minneapolis, and Martin in the back of a van with his artistically tempered anarchist band, Creation is Crucifixion. And without that punk background -- and, particularly, the do-it-yourself, nothing-is-impossible ethos that went with it -- Maphub could never have gotten off the ground.
"I saw one of those horrible-yet-captivating documentaries on the history of punk rock the other night," says Carl DiSalvo, "and they were saying what they always say: 'You didn't need to know how to play guitar, or to run a studio, to form a band or to cut a record.' And in a sense, that's exactly what we did with Maphub. We'd never even touched mapping software before, but we just said, 'Technology's not scary, we can do this.' It totally fits with our prior lives."
While working together at MetaDesign in San Francisco, Martin, DiSalvo and colleague Hans Meyer (now a geography Ph.D. candidate at Cornell) began brainstorming three related ideas. One was a solution to a simple but very real problem: In a superheated housing market like the Bay Area, by the time an apartment-hunter found out about an available apartment, it was gone. Why couldn't information about available apartments be shared more quickly, and in a geographically based way?
With an online map, with icons denoting available spaces that could be updated regularly by any one among a group of people, someone could look at the pertinent area of a neighborhood and see places actually available -- not listings of apartments that had been spoken for days before. Apartment-hunting didn't capture DiSalvo, Martin and Meyer's attention for very long. But the implications of the solution they conceived quickly became apparent.
As Martin recalls, "We started thinking, 'OK, so we can spatialize information'" -- take information and pin it to a specific location on a map. "'So, who is that [ability] important to?' If you look back historically, the most important group looking to spatialize information is the military. And any military project can easily be subverted for use by protesters."
The resultant idea was conceived in 2003 -- not coincidentally the year of massive anti-Iraq invasion protests. The concept was Maptivist: a map-based communication tool to enable protesters to relay information before, during and after mass protests. Maptivist would've allowed its users to plan protest activity in advance, and to monitor police and protester movements during a demonstration, in order to plot escape routes and evaluate the group's best moves for impact and safety.
Martin and DiSalvo took Maptivist around the academic and tactical-media communities, showing off their idea at such influential events as Chicago's Versionfest and University of California at San Diego's Infrastructures of Digital Design conference. But despite their academic success, Maptivist was no more than an idea -- a set of written documents, schematics, Power Point presentations and mock-ups demonstrating what could be done. Martin and DiSalvo's expertise was in design and art, not computer programming.
But as their thoughts expanded outside of the narrow world of protest politics, they wondered: If protesters could use this project to share timely knowledge about city streets, why not take it a step further, and allow anyone to share knowledge about his or her community?
In 2004, Martin was awarded an artist-in-residency at Carnegie Mellon's STUDIO for Creative Inquiry to work on the group's mapping project. DiSalvo entered a Ph.D. program at the university, and both moved to Pittsburgh. Once here, they recruited two CMU grads: Evan Merz, a computer programmer, and Jeff Maki, a recent graduate of CMU's Information Systems master's program
Maki and Mertz, the team hoped, would provide the technical expertise necessary to implement DiSalvo and Martin's ideas. But by this time, their concept had changed again.
What it changed into was Maphub -- an easy-to-use interface for people to map and share information about Pittsburgh. To Martin, this change was in line with the philosophy that Carbon Defense League had always espoused -- the idea that political change doesn't come from some kind of spontaneous revolution, but from slowly changing the way people see, and access, their surroundings.
As Martin explains, the idea is to "effect change on small behaviors." Such an approach, he says, can have more lasting effects that creating "some kind of uproar and then have that change trickle down, which is the idea of traditional protest."
"One of the things unique to [Maphub] is that they really want to work within communities," says Kristen Kurland, Geographic Information Systems (GIS) specialist at Carnegie Mellon. Kurland works with CMU's School of Architecture to teach students programs such as ARC GIS, the most widely used map-making software package. (It is, in fact, the system with which Maphub's base maps are created.)
Kurland is working on a book about GIS and health, including a project which maps correlations between lead levels in homes and children's health. Systems like ARC GIS "are great, but they're of somewhat limited use," she says. It takes an entire CMU course to learn the software -- assuming you already have a background in programming.
"[Y]ou [normally] need an expert user to map these things out," Kurland says. As a result, "From an urban-design perspective, we usually work with the community, and then go back and map out the information that they give us." But in the real world, conditions are constantly changing, and "data becomes stale if you don't actively update it. If we could have citizens more involved in [sharing the information directly], you get more timely, accurate information."
That's where Martin and DiSalvo come in.
"Where Maphub is unique is it allows regular people to upload information," says Kurland: The program is simple enough that the average Internet user can be inputting data within an hour. Maphub, Kurland says, isn't "a software application, but a participatory tool that people can improve their community with."
Martin and DiSalvo themselves, in fact, lacked the computer skills they needed to accomplish their goals. And that lack of knowledge, they agree, probably helped Maphub to happen in the first place.
"When you're invested in technology, and start thinking about projects you might want to undertake, you think to yourself, 'That's going to be so difficult to implement,'" says Merz. Maphub, after all, takes the already complicated ARC GIS program and couples it with an online component. It incorporates mapping information from a myriad of sources, including census and city planning files. And it allows multiple users to constantly update the map and database simply and quickly.
"When it comes to programming the Web site, [DiSalvo and Martin] will make demands that are ridiculous," says Merz. But those demands are also what make Maphub a viable community tool.
Or it will be, once with community finds out about it.
The STUDIO for Creative Inquiry was founded in 1989 to connect the arts to other disciplines prevalent at CMU, such as robotics, biology and ecology. But STUDIO work often flummoxes those outside the art world -- and for that matter, some of those within it. Nathan Martin, for example, started his work as an intern on a project which involved a performance artist wearing a robotic arm.
Marge Myers, associate director for the STUDIO, understands all too well that in the no-man's-land between art and science, or art and community development, audiences on both sides can end up feeling alienated.
"A couple of questions come up in a lot of the projects generated here," says Myers. "And one of them early on is, 'Who cares about this?' Often it's not a lot of people, and in a way I think that was true [of Maphub]. There were technical and media [applications] to mapping using the Internet, but not for the purposes of helping everyday people.
"With some of the work created here, we're not sure what's the right vehicle to let people know, so that's part of the challenge -- finding the people, getting them ready to experience it."
Since the launch of Google Maps -- the search-engine giant's online mapping tool -- the cultural craze for so-called "locative media" projects has reached a kind of frenzy. Just look at the blog Google Maps Mania (http://googlemapsmania.blogspot.com/), which catalogs Web sites using Google Maps to chart everything from great places to get a beer to where you'd come out if you dug straight through the earth from a specified location.
What makes Maphub unique, however, is its designers' dedication to working offline more than online. Maphub is meant to be used by people working toward a community goal -- rather than a lone blogger whose site is the beginning and end of the project.
"Locative media is too often about the media, and not about the location," says DiSalvo. "[I]f you really want to do this work, it's about relationships, building relationships with groups, that don't happen over e-mail or overnight. It's about knowing things about the city that you only know because you've walked down these streets a thousand times."
One of the early projects undertaken by Maphub, for example, was an art piece called HeardHub -- an installation which allows viewers to play back, directly into Maphub, sounds recorded via cellphone in various locations around Pittsburgh. Open the hub, and icons locate where sounds were recorded. Click on an icon and you'll find an MP3 file of the sound. These range from unrecognizable, ambient traffic sounds to electronic noise compositions, or an anonymous person ordering in a coffee shop. HeardHub was featured at the ZKM Museum in Karlsruhe, Germany; recently, Maphub programmer Evan Merz created a computer program which allows him to perform compositions using the Heardhub sounds. (Merz is a full-time piano instructor and musician, and perhaps the only person ever to use music as a financial backup to his CMU computer degree.)
A similar Maphub project is designed for the Pittsburgh Signs Project, which has linked photographs of iconic regional signage onto the Maphub site. But while such projects are of great interest to Maphub's creators, these aren't the uses which they are so desperate to land.
"We've made some projects with it that are artistic and ephemeral," says Martin, "but Maphub itself, no, I don't think [that's what it is]. And that's what's really been demanding on us, because now we have all these responsibilities that artists don't normally have to worry about. 'You remember those community groups we said we were gonna work with? Now we really have to do it.'"
The Bike Crash hub is one example of how Maphub can impact the community's health, and a similar opportunity may be close at hand. One of Kristen Kurland's projects is a collaboration with insurance giant Highmark. Its goal: addressing childhood obesity by getting people more physically active. Highmark is discussing using Maphub technology to make real-time maps of information regarding amenities, sites of interest, and activities to get people out of the house. It's just the kind of micro-behavioral change that DiSalvo and Martin have been working toward.
But it hasn't come easily. Other potential projects -- including one to work with Pittsburgh City Planning so that Maphub-based input can be used in zoning hearings -- have stalled for reasons beyond anyone's control. (A new crop of mayoral appointees in city government, for example, means Maphub has to make its case to different administrators.)
"They're not out to sell the tool," says Kurland, "and that's where it's a little more difficult for them. ... If they wanted to just sell the tool, they probably could and could make a lot of money, but that's never been their goal."
It may end up happening anyway.
Jeff Maki looks uncomfortable in his clothes. As he sits at a table in the STUDIO for Creative Inquiry, entering more computer code, he wears his tucked-in button-down shirt and black slacks the way a previous generation of first-time office workers might've borne a necktie and suit jacket.
Maki is waiting for Martin -- himself on his way from his first two jobs, coffee and dinner in hand. DiSalvo, who is working on the finishing touches of his doctoral thesis for CMU's school of design, is also running late: Presumably he's taking a few minutes to check in with his wife and child. Evan Merz, whose paid position programming for Maphub ended months ago, continues to work on the project whenever time allows.
"We're in a position where we can only dedicate 15 or 20 hours a week [to Maphub] between all of us," says Martin. "And right now it's tough to even do that without funding."
"Jeff is no longer a student," says DiSalvo. "I'm no longer going to be a student within six months. Nathan hasn't been for a couple of years. Still, I think there's a desire more than a need to get some funding for some of these projects -- a desire to [be able to] continue pushing our work in new directions."
One possible direction points toward Rob Conway, a licensing officer at the Carnegie Mellon Center for Technology Transfer (CTT). The Center is charged with helping campus-funded technology developers find companies to invest in their ideas. A tool such as Maphub, for example, could be put to a myriad of commercial ends: use by the travel industry, say, or by an Internet firm like Google or Yahoo for use in a larger application.
Since Maphub was largely developed while Nathan Martin was a CMU artist-in-resident and DiSalvo and Maki were STUDIO fellows, CMU can claim some ownership of intellectual property in Maphub. The Center could license the program for commercial use, with money being split between the inventors and the university. If, that is, Rob Conway can find an interested party.
"Hopefully [one of those companies] will want a license, but it's a long process of determining interest," Conway says. "We're just at the marketing stage at this point."
Another option for Maphub would be to spin a small company out of the university in order to develop commercial applications. In that case, the Maphub team itself would be given the license, and would develop other products using the technology. It's a fairly regular method of moving technologies along: CTT launched seven start-ups in 2005 alone. But it's rare to find a spin-off company with origins in an art project.
Still, says Carl DiSalvo, "I don't see [the commercial and community development sides to Maphub] as being separate at all. In some ways, to have the greatest impact it needs to become formalized." He admits, however, that, "I'm not sure it's what I want to do right now. It's all-consuming, and at the end of the day, I don't know how interested I am in developing technology except for the ends that it might provide." What's most important, he says, is offering a chance for "dialogue between different groups. ... Whether it's Maphub or a burlap sack we use [for that] doesn't matter to me."
Regardless of who may license Maphub's technology, its inventors will retain the right to use it themselves. The existing Maphub.com, with its focus on Pittsburgh, can go on as before, adding new hubs where there is community interest. But selling the technology may also free Maphub team members to start exploring uncharted territory themselves.
The opportunities are enough to get DiSalvo, Martin and Maki excited about sitting in a conference room, tired from 12-hour work days, talking about potholes. Not in the post-commute-whining kind of way, but with a grave artistic seriousness -- the way others might discuss "Guernica." The idea they're exploring involves fitting a laser range-finder -- like a massive computer image-scanner -- on the bottom of a car, then driving around cities, creating visual artworks out of maps of their pockmarked streets.
Afterwards, DiSalvo sees the humor in it all.
"When was the last time you heard people talk so much about making fucking art about potholes?"
It's one of a half-dozen ideas DiSalvo, Martin and Maki have thrown around as something they could present outside Pittsburgh. Despite its slow integration into use by Pittsburgh's community-development organizations, Maphub has garnered kudos from around the locative-media arts world, earning its creative team invitations to participate in arts conferences and competitions across the country. The challenge is to allow Maphub to "travel" by finding new uses for the technology -- or new ideas based on the same ideology.
With the possibility of revenue from licensing their invention, or from a funded project with an organization such as Highmark, comes another possibility: dedicating more than the 15 or 20 hours per week the team has had for Maphub since Martin and Maki took on full-time jobs at Downtown tech firms. What to do with that time is a question they're just starting to address.
"I'm a 'technical' person," says Jeff Maki, "and a lot of people would say that because I can't paint -- because writing code is what I'm good at -- I can't make art. But [Martin and DiSalvo] say that I can, because I do have something to say."
"What's a great opportunity here is, 'What are the other tools we can create to foster this kind of communication, even staying with mapping?'" says DiSalvo.
One idea they've been developing is Trailposts.com, which Martin hopes the team will launch by March 1. Trailposts.com is a Maphub-like interactive map of the Appalachian Trail, the 2,000-mile long hiking trail that runs from Georgia to Maine. While hiking parts of the trail last summer, Martin discovered what so many hikers have before: Information about the trail found in books and maps is often outdated. Where there was water the previous season, there may not be today. Shelters are damaged, civilization encroaches, opportunities arise.
With hikers adding information to an online Trail "hub" on their weekly or bi-weekly re-supply trips into nearby towns, Trailposts could act as a constant source of information -- much like the Bike Crash hub serves Pittsburgh cyclists.
It's not Pittsburgh, and not "community" in the way we normally think of it: The Trail's approximately 6,000 hikers each year come from all over the country and will most likely never see each other. But it's in keeping with the Maphub philosophy: creating computer-based geographic tools that take people off their computers and into the real world.
"One of the interesting questions about Maphub is, 'When is it ever over?'" says DiSalvo. "I have a daughter who was born the same time we began [discussing this project], so my daughter is the same age as Maphub. And now, she's 5-and-a-half. We came up with this not as an accident, but also not as a plan, and that's what keeps our work interesting.
"So the thing I'm always thinking is, 'At what point do we allow the next accident to happen?'"