Seeing like a Rover

How do you work with a robot millions of miles away to make scientific discoveries on a planet you have never set foot on? While the team of NASA scientists and engineers rely on Spirit and Opportunity to do science on Mars, working with the Mars Exploration Rovers also requires that team members learn to "see like a Rover". On the one hand, this means developing visual fluency with the hundreds of thousands of images that return from the Martian surface, producing new representations with image processing software to inform Rover operations. On the other hand, this visual connection to the Rovers' "eyes" on another planet produces a deeper connection to the robots too: one that ascribes human characteristics to machines, teaches humans to see, move and feel like their robots do, and develops an intimate and embodied understanding of the vehicles' experiences on Mars. All this takes place among team members collaborating across vast distances on Earth, from JPL to research laboratories across the US and Europe.

Now published as a book by the University of Chicago Press, my two-year ethnography of the Mars Exploration Rover team was fueled by an interest in digital images in science: how scientists work with them to produce scientific analyses; how manipulated images are seen as credible, trustworthy and scientific; and how they structure and inform interactions between team members and their robots. My dissertation focuses on digital image work to explore how images are active sites of work and interaction among a group of scientists, who together have to learn to "see like a Rover" to do science on Mars. My first book project picks up on these questions and aims to show the relationship between digital image processing - a seemingly singular activity conducted alone at one's desktop computer - and social order. I develop and deploy the concept drawing as, a new analytical framework to better understand the relationship between scientific representation and observation, and to demonstrate how visualization techniques produce and restrict certain ways of knowing. I then place the work of purposeful image craft and manipulation in the context of the Rover team's work on Mars, showing how and why the team's social organization matters to the interpretation of images and the execution of robotic activities on Mars.

Seeing Like a Rover: How robots, teams and images craft knowledge of Mars is now available from your favorite booksellet. Several of my publications address this issue in a shortened format: for instance, this piece in Contexts Magazine or my award-winning article in Social Studies of Science.

Back to Research