Pulse Logic transforms the threshold of Nine Elms Underground Station into an adaptive civic infrastructure. The project begins with a custom live interactive map. Since the only data TfL makes public is a limited "arriving in N minutes" countdown — not live vehicle positions — the project works backward from these values, using each station's and stop's average arrival and travel times to reconstruct the estimated positions of trains and buses along their route geometries.

Run over several weeks, this map accumulates its own dataset, from which the project derives the Arrival Wave Index (AWI) — a predictive measure of how much crowd pressure is building toward the station. The AWI lets the architecture act pre-emptively: rather than reacting after an arrival wave has landed, it reads the approaching pressure and responds in advance.

This becomes the basis for simulation across ordinary, rush-hour, disruption, and event conditions. Each simulation models how collective pressures — arrival waves, queues, dwell, heat, points of congestion — form and move across the site, and tests alternative architectural responses to them. A subsequent real-time section reconnects the same system to live data, allowing apertures, louvres, lighting, ventilation, and shared waiting spaces to change with current conditions.

At no stage does Pulse Logic track individuals; it works only with aggregated, predicted data. Instead of using AI to police or optimise movement, it makes collective urban pressure publicly legible — and leaves open a final question: who should govern this adaptive infrastructure, the AI that operates it, or the people who read and respond to it?