Currents Atlanta, GA 2017
Commssioned by IBM

As visitors move through the lobby of IBM Watson’s headquarters in Atlanta, light animates across a folded wall shaped by wind patterns and driven by real-time air quality data. Shifts in color and motion translate invisible atmospheric conditions into a spatial experience that unfolds through movement and perception.

Currents is an interactive wall that visualizes global air quality using live data from The Weather Channel’s Air Quality API. Through an interface embedded directly into the wall, visitors can explore cities with low, high, or rapidly changing air quality. Each selection reshapes the installation, converting Air Quality Index values into flowing vectors that animate the surface through light.

The wall is composed of a tiled field of folded aluminum panels, each angled to suggest currents of motion inspired by atmospheric flow. These folds shape how light is reflected and diffused across the surface. Behind each panel, a grid of LEDs generates color values derived from averaged air quality data, softened by diffusers and white powder-coated aluminum surfaces.

As viewers shift position, the orientation of the panels reveals the illuminated grid differently from every angle. This lenticular effect creates variable hotspots and moments of emphasis, introducing an analog quality that works alongside the digital visualization. Light appears to pool and drift across the wall, giving the installation depth beyond that of a conventional screen.

Through the combined behavior of folded geometry, diffused illumination, and live data, Currents gives light an almost material presence. Rather than presenting information as static content, the installation allows environmental data to behave spatially—revealing global systems through motion, texture, and change.

Photos: Alan Tansey

Find out more about the process:
Weighted average
Photos of testing and installation

Currents - SOFTlab 2017
Currents - SOFTlab 2017
Currents - SOFTlab 2017
Currents - SOFTlab 2017
Currents - SOFTlab 2017