CJ KWEON / PULSE LOGICBack to project ↗

FROM DATA TO MOVEMENT

A responsive fabric.

Rain lifts the fabric above its hinge. Sunlight opens a different posture.
Arrival activity continues as a quieter movement within that posture.

Loading the Rhino geometry…
0.0°0.0°0.0°
0.00.0
0.00.0
0.0°

Explore · synthetic inputs, not live measurements.

Short studies accelerate weather transitions for comparison. Live transitions ease over approximately one minute. Proposed postures; weather protection is not yet validated.

Weather sets the posture.

Rain raises the free edge above the hinge, with less twisting. Water paths follow the model’s actual downhill slope toward the root or other low edges; amber rings mark possible local traps. A proposed root collector and side drains are shown. Strong sunlight favours a more open posture with a smaller travelling fold. When both occur, rain takes priority. Arrival activity adds a call, a brief hold and a reply within the weather posture.

One movement, two views.

The right view exposes the same central rib. Thin red and blue lines brighten when their drums are commanded to wind in. Numbers show upper bend, lower bend and tip twist (degrees), paired tendon take-up (millimetres), and whole-panel lift. Brightness does not measure tension.

Data, method & session export

Live reads the existing map AWI producer at 51.479912, −0.128476. Raw AWI is unbounded; the map display is min(9.9, raw × 0.42). This study maps raw AWI to 1 − exp(−raw / 12) for motion. This new design mapping is separate from the earlier section response index. TfL inputs are arrival forecasts, not measured passenger counts.

Weather comes from the Open-Meteo weather model ↗ at the same requested point. The current model interval is fifteen minutes. Current rain and showers are converted from millimetres per interval to mm/h. Solar radiation is horizontal shortwave radiation, not radiation measured on this fabric. Weather is polled every five minutes (every thirty seconds after failure); TfL receipts update about every twenty seconds. Source and receipt times are shown in UTC.

Rain enters at 0.4 mm/h and exits at 0.1 mm/h; sunlight enters at 600 W/m² and exits below 450 W/m². These are provisional design thresholds. Missing weather holds the last pose. Missing AWI suspends new arrival gestures while a valid rain or sun response continues. Gusts at or above 12 m/s hold the current pose; this only stops simulated motors; weather traces continue, and the raised pose is not a certified wind-safe position. Short studies use synthetic inputs and a 2-second smoothing constant; live uses 25 seconds.

Weather overlays are geometric studies: vertical authored rain, downhill traces and possible local traps, not a fluid-volume or drainage-capacity calculation. The membrane is assumed impermeable. Dim water traces are shown through the translucent cloth for legibility; brighter moving marks are illustrative tracers. The main view includes the proposed collector; the inset isolates the original rib mechanism. Collector sealing, flexible connections and hydraulic capacity remain unresolved. Drip paths omit contact with building and drive housings, including during the lift transition; clearance has only been checked at the final rain pose. Light rays use an authored direction and an opaque interception footprint on a reference plane at z = 0, not a measured solar orientation or calibrated fabric transmission. Main AXO is unchanged; the mechanism inset keeps its camera direction and position, with a wider fixed extent to show the raised rib.

The cloth is ported from the Rhino study: two constant-curvature rib sections, antagonistic tendons and a tip vane. Deep folds preserve an assumed cut width; intermediate twist is an assumed drape. Gravity, contact, wind, motor loads, fabric stress, drainage and shading performance have not been solved. Site north, material permeability and structural limits require confirmation. This page sends no motor commands.

Rendered commands and input provenance · up to 3,000 samples