Installation Showing the Perspective of a Self-Driving Car Aims to Evoke Empathy for Artificial Intelligence


Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

Driver Less Vision, an installation at the 2017 Seoul Biennale of Architecture and Urbanism by Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin, is an immersive 3D video experience comprised of spatial scans of Seoul, projected into a dome and paired with surround sound. The supporting audio is the internal monologue of a personified autonomous vehicle, driving through the streets of a future Seoul, Korea. The installation transports vierers to the front seat of the autonomous vehicle, providing a new perspective of traversing cities—through the car’s point of view.

Driver Less Vision was a response to the Seoul Biennale’s theme of “Imminent Commons,” which asks questions about what will become of cities and the role they play for both the individual and the collective. The installation takes recent discussions of what autonomous vehicles will mean for society and combines it with the potential technology has for interacting with humans. Driver Less Vision explores some of the urban obstructions an autonomous vehicle might encounter to reveal ways this can affect autonomous mobility and the future built environment.


Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

At the Biennale, the installation is staged inside an 8-meter (26-foot) diameter dome suspended from the ceiling. Attendees will duck their head as they enter the space, becoming immersed in the projection and all-encompassing sound. The three designers, Urtzi Grau, Guillermo Fernández-Abascal, Daniel Perlin, developed the 360 degree experience along with the University of Technology Sydney, Rice University and Ocular Robotics.


Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

In the monologue, the autonomous protagonist speaks in a way that suggests machine learning, a computer science subfield that gives computers the ability to learn without being explicitly programmed. This is increasingly a subject that requires urgent investigation: software is already implementing machine learning, including programs that register the surroundings of autonomous vehicles and make them capable of responding to unpredictable errors, such as drunk drivers.


Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

The Driver Less Vision installation brings us to question how we will design autonomous vehicles in the built environment. There are many companies racing to get autonomous vehicles on our roads. Legislative bodies and individuals are gradually taking their stances on what this new technology will bring to our daily lives. In many ways, self-driving vehicles are a very attractive proposition for cities, given the potential for improved road safety, the ability to reduce parking requirements or move parking outside of city centers, and the reduction in traffic.


Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

Courtesy of Guillermo Fernandez-Abascal, Urtzi Grau and Daniel Perlin

But to what degree do we want to incorporate this new technology into our city master plans? Projected urban migration patterns, the exponential rate of innovation in current technologies, and the role government agencies will take all fuel the question the Seoul Biennale asks. Seung H-Sang, Chair of the Biennale’s Steering Board asks, “what is a good city? […] The space and structure of cities, creative development and regeneration, new building methods and technology, sustainable urban environments, city governance, and new forms of solidarity: these are the central issues of the contemporary urban generation.” The Driver Less Vision installation helps us to explore these questions and potentials.