road-like curb lines. We then calculated the
average vanishing point of these lines, and
as long as we had a clear and open road with
a painted curb and few shadows, the road
detection system worked well. The system
also worked while looking down hallways, and
in unexpected situations, like making a pyramid shape with your arms above your head.
Although the vision system had several
inaccuracies, it was considerably more
flexible and playful than the location-based
GPS system. We built our own version of the
original arcade driving game in Flash, and
used our calculated computer-vision road
endpoint to change the shape of our Flash-based game road (Figure E).
VIRTUAL REALITY ON WHEELS
Making physical things is often a battle with
tools and materials, and this project was
complicated by trying to integrate computer
vision, physical computing, custom software,
and other systems — like a golf cart.
However, actually making a working real-world system has a significant power of legibility
and reality over just mocking up something in
Photoshop. Rolling down the street, the arcade
cabinet car is understandable and real to a
diverse audience — grandmothers and infants,
for example — in a way that my Photoshopped
image wasn’t. And the process of wrestling
between the materials and the concept had its
own imperfect, wabi-sabi beauty to it.
The OutRun project was never intended
to seamlessly turn a video game into real life
and real life into a video game. Instead, it was
intentionally built as a type of chindogu or
critical design — a system that uses paradox,
irony, and physical prototyping to raise a
series of questions in a provocative way.
In the case of OutRun, when the game is
extended beyond its normal constraints as
a video game, its playfulness malfunctions:
it’s like a fantasy taken too far, and it results in
confusion, nervous humor, and a questionable
sense of over-reliance on technology.
To build in this way is to slip into the role of
a trickster: using humor and paradox to challenge, bend, and break common assumptions
about our everyday lives with technology.
Videos and build notes: conceptlab.com/
Thanks to Walt Scacchi, Chris Guevara, Alex Szeto, Paul Dourish,
Gillian Hayes, Jong Weon Lee, Eric Mesple, Yuzo Kanomata,
Jeremy Bailey, David Dinh, Matt Shigekawa, Jesse Joseph, Mike
Tang, Matt Wong, Erik Olson, Richard Vu, Kari Nies, and Craig
Brown for their development efforts and assistance.
Support for this project was provided by the following organizations at the University of California Irvine: the Laboratory for
Ubiquitous Computing and Interaction, the Arts Computation
Engineering Program, the Center for Computer Games and
Virtual Worlds, the Institute for Software Research, and the
California Institute for Telecommunications and Information
Technology. Support for this research is also provided through
grants from the National Science Foundation (#0808783)
and the Digital Industry Promotion Agency, Global Research
and Development Collaboration Center, Daegu, South Korea.
No review, approval, or endorsement implied.
Garnet Hertz is a research scientist and artist-in-residence
in the Department of Informatics at the University of
California Irvine and is adjunct faculty in the Media Design
Program at the Art Center College of Design in Pasadena,
53 Follow us @make