Drifting Atlases

Overview

Drifting Atlases proposes a kind of map that is never finished — one that learns the shape of a city from how people actually move through it. Instead of representing streets, it represents habits, hesitations, and shortcuts, growing a softer geography of a place.

Research Objectives

• Build real-time sensing that reads movement without identifying people
• Let the atlas rewrite itself as the city's routines shift
• Test spatial computing as a medium for collective, debatable memory
• Design interfaces where the map is something you tend, not read

Methodology

We combine lightweight motion sensing, generative cartography, and co-design workshops. Anonymous traces of movement are distilled into drifting regions that participants then argue with, prune, and rename — the atlas becomes a negotiation rather than a measurement.

Key Findings

Early fictional deployments show that people trust a map they helped shape far more than one handed to them. Participants formed attachment to regions that never existed on any official plan, and used the atlas to imagine infrastructures the city had forgotten.

Future Directions

Next we plan to let multiple cities' atlases drift into one another, add a gentle haptic layer for wayfinding, and open the authoring tools so communities can keep their maps alive after the residency ends.