Windows on Last Miles
Mapping city space with interactive art
Project Narrative
Figure 1. Sketch of possible installations and photo from early installation test (15. May 2008)
One of the uses of the Internet, if not its defining use, has been to connect people separated by physical geography. Can the Internet connect people who are geographically close but separated by other forms of distance? The project, Windows on Last Miles, aims to investigate and perhaps bridge (if only momentarily) distances other than the geographical. The installation consists of two mobile installations, a computer display (the "window") and an interface, which connect two nearby public spaces in real time. The installations will be placed in pairs of sites (Figure 2), selected because they are "distant" in ways other than the geographic—they are, in all cases, a mile or two apart. Via the installations, these "distant" public spaces become connected, visually and aurally by the two "windows," which relay feeds from two adjacent web cameras. These windows are symmetrical and interactive. Each participant sees and hears what is happening in the remote location and can control the remote web camera, panning and tilting it with a joystick interface in order to explore the remote location. Examples of connected public spaces might include the courtyards of Duke University's Perkins Library and the Durham City Library, or a municipal bus station and the university bus circle.
The windows are mobile. Requiring only an internet connection and power, the windows are intended to be left in place for a short period of time (perhaps a few days) before being moved to other pairs of public spaces. The hope (and hypothesis) is that the observed use in the various pairs of public spaces will reveal the “connectedness” of these distant-yet-proximate places. The information gained, we hope, will allow us to add a sense of a social and cultural geography to the physical. A range of measurements will attempt to capture the differences in use in these different pairs of spaces. The end result will be a kind of cartogram that expresses the findings of the project, representing city geography in a novel and hopefully useful way (see Figure 3 for an example of a cartogram attempting to capture population density alongside geography).
The windows are also unannounced. While intended to be highly visible once in place, they are simply open to interaction and experimentation by visitors and passersby. Equally unannounced will be the possibility for users to communicate with each other via attached headphones and microphones. Two-way communication is intended not as an afterthought but rather in anticipation of and hope for unexpected connections, organization, and collaboration. What might have been done as a survey, showing passersby images of the remote site, becomes something that invites the unanticipated.
The first installation in public spaces is planned for the beginning of the 2008-2009 academic year.
Project Goals
Mapping Social Geography
What might be done with the data generated from the experiment? Could one begin to construct a window onto the social geography that contains these public spaces? Knowing that two spaces are distant (again, in a sense other than the geographical) but that users are passionately interested (i.e. frequent, intense use of the installation) in what is happening in the other public space seems like an important result.
The following are some of the measurements that could be collected unobtrusively and automatically: users per hour, user feedback (gently interrupting the user of the window periodically with survey questions like “Have you visited the remote location in the last month? (six months, year, ever)”) (See Figure 4).
The pairs of locations that might be connected include:
| Duke East Campus | Duke West Campus (as a control of sorts) |
| Nasher Museum | Museum of Life and Science |
| Duke Library | Durham Main Library |
| Duke West Campus Bus Circle | Durham DATA Bus Station |
| Duke West/East Campus | Durham Public Middle School |
| NCCU Quad | Duke Quad |
| UNC Chapel Hill Quad | Duke West Campus |
| Exit of Griffith Theater (Bryan Center) (Duke) | Exit of Carolina Theater (Durham) |
| Cameron Stadium (Duke Basketball) | Durham Bulls Stadium |
Investigating Telepresence
The windows are also unannounced. While intended to be highly visible once in place, they are simply open to interaction and experimentation by visitors and passersby. Equally unannounced will be the possibility for users to communicate with each other via attached headphones and microphones. Two-way communication is intended not as an afterthought but rather in anticipation of and hope for unexpected connections, organization, and collaboration. What might have been done as a survey, showing passersby images of the remote site, becomes something that invites the unanticipated engagement of users.
A larger question concerns telepresence and teleconferencing generally. Teleconferencing software has been used principally for connecting people who were already connected. For example, software is used to connect rooms at Yahoo Mountain View and Yahoo NYC. Telepresence in this case serves to connect people who are already connected. Could teleconferencing devices and software be re-purposed for other ends? Perhaps to connect people who have never been connected. Is there any interest in communicating directly with unfamiliar people? Does the context of an art installation (an instructional sign will identify the project as an interactive art project) make a difference?
Promoting Interactive Art
With an eye to a possible future of student and faculty creative work in the area of visual, digital, and interactive art and design, every effort is being made to document the project and to make use of accessible open source software and hardware. All code for the project will be released under an open source license.
For example, the user input via the joystick and the pan-tilt setup underneath the webcam is facilitated by an Arduino board. Arduino is a simple I/O board and development environment that's licensed under the GPL and LGPL. The question interface and the relaying of joystick commands over the internet to the remote site is managed by a small Python program, using the open-source GTK toolkit and Python standard libraries.
By documenting every step of the project online in a production blog, we hope to give others the tools and inspiration to begin their own interactive art and research projects at Duke.
Project Timeline
2008 January: begin development and testing.
2008 late March: first live test. Connect Duke East and West Campus bus stops. Continue gathering permissions for May installations.
2008 May: First test.
2008 August-September: Public installations as weather permits. Connect Duke and Durham Library, Nasher and Museum of Science. Duke West and NCCU, etc. All pending permission.
Project Team
Christina Patsiokas
Brenna Heitzman
Quentin Conrate
Acknowledgments
This project benefited greatly from the support of Tim Lenoir and the Information Science + Information Studies (ISIS) program. Victoria Szabo was particularly instrumental in seeing it through to this stage. HASTAC and Duke's University Scholars Program provided conference support. Many thanks as well to Sarah Wilson, Katrina Poetzl, Haun Saussy and Mel Chin, Molly Rowe, and Victoria Lodewick. Thanks to Bill Price and David Berberian for carpentry consulting.