New York City's subway turnstiles are pushed millions of times each day. Every rotation of the metal arm is logged by the MTA's system — timestamp, station, direction, count. The original purpose of this data is fare management and operational scheduling. Nobody designing the turnstile mechanism was thinking about art. The data is the sediment of commuting behavior, a byproduct left behind after the transit system does its job — in SAE terms, a vast remainder archive. Trpti Sanghvi's Turnstile Murmurations, shown at the Data Through Design 2026 exhibition "Echo(logies)" at Brooklyn's BRIC House in March, does something structurally precise with this archive: she feeds it into another remainder-generating mechanism — the murmuration (starling flocking) algorithm.

Murmurations are the textbook case of emergence. No individual starling "designs" the pattern. Those breathtaking aerial formations are the residue of each bird following simple local rules — maintain distance from neighbors, align direction, move toward the center. The pattern is the remainder of the rules, not their goal. What Sanghvi constructs is a double-remainder stack: turnstile data is the remainder of commuting; the murmuration pattern is the remainder of agent rules. When the temporal rhythms of millions of turnstile rotations are mapped as behavioral parameters for a virtual flock, the shapes that appear on screen belong to neither the vocabulary of transit planning nor the vocabulary of biology. They are what is left after two systems' chisel-construct cycles collide — something no one planned for.

There is a critical structural observation here: Sanghvi is not a "data artist" in any conventional sense, nor a biological simulation programmer. Her background spans Odissi classical dance, data analytics (her day job is product analyst), neuroscience interests, and algorithmic art. On her personal site, she describes her art as "abstract expressionism in the digital age, with the body acting as a paint brush on the canvas of the internet; the memory of movement is captured with color." This means she approaches data flows from the perspective of the body and of dance — the pushing of a turnstile arm is itself an accumulated bodily gesture, closer to choreography than to a spreadsheet. This perspective is not taught by any existing discipline. It is the remainder of her particular lived experience.

Seeing this work now matters more than seeing it later, because it is still in its remainder state. Data art as a category has existed for two decades, but its dominant grammar remains an extension of information visualization — beautified charts, poeticized dashboards. The flocking algorithm is already-construct in the biological simulation field. But feeding urban transit data into a flocking algorithm and letting emergent patterns reveal hidden rhythms in collective human behavior — this operation does not yet have a category name. It is not "data visualization," because information delivery is not its purpose. It is not "bioart," because no actual organisms are involved. It is not "urban art," because it does not happen on the street. It sits in the gaps between these terms — still a remainder, still alive.

Han Qin (秦汉) · @hqinjarsy · SAE Judgment & Aesthetics