Heat, hydrology and ecology, read as one system
B6 is not measured as three separate datasets but as a single coupled field. Air and surface temperature, humidity and soil moisture, movement and dwell are read against each other — because in a courtyard they are not independent. Where the ground overheats, evaporative buffering collapses; where buffering collapses, people leave. The aura above renders that coupling continuously, rather than as a snapshot.
Across the B6 dataset, air temperature and occupancy move in clear opposition — a correlation of r(T, occupancy) ≈ −0.72. The hottest hours are the emptiest hours. The field makes that relationship legible in space, not just in a chart.
envelope −1.6 / +47.1°C Array min–max · S1–S4
point (MKT) S2 · 19.3°C East facade
point (mean) S3 · 13.2°C West facade
humidity 10 – 98% Full measured range
readings 40,070 10-min interval · 4 pts
window 28 Mar – 6 Jun 2026 · ~69 days, ongoing
Four points, two hot poles, one cool corner
Phase I instruments the courtyard with four surveyed sensing points, logging air temperature and humidity continuously at a 10-minute interval — one layer of a wider array. The latest reports (28 March – 6 June 2026, ~40,000 readings) describe a field with clear structure: the sun-exposed entrance and balcony, S2 (east) and S4 (west), form two hot poles — peaking at +47.1° and +43.7° and carrying the highest sustained load (MKT 19.3° / 19.1°). The shaded garden corner, S3 (south), stays the coolest sustained point and holds the deepest minimum at −0.8°. S1, in the street, runs as a separate reference climate.
A central basin that overheats, edges that stay live
The reading is consistent across the diurnal cycle: the central courtyard basin accumulates heat through the afternoon and tips out of usable range after roughly 13:30. High surface temperatures there correlate with low dwell — the centre empties as it heats. The edges, cooler and better buffered, stay occupied. The imbalance is spatial, and it is the imbalance the interventions are designed to correct.
This is diagnosis before prescription. The field measures the real condition first — then the intervention packages are tested against it, live, in the same instrument.
Three intervention packages, tested in the field
Each package is a coordinated set of spatial moves aimed at the measured condition — not a generic retrofit. Preview any of them in the live field to see the projected change to the heat surface.
Tip — preview a package live: open Tweaks and switch the field demo, or append ?demo=IP-02 to the instrument.
Built to receive the live edge feed
Today the field runs a realistic 24-hour diurnal simulation. When the Ionity edge bus comes online, readings push straight into the running instrument through a single integration point — the simulation gives way to live conditions with no change to the interface. Edge-first, locally stored, exportable: AEDI does interpretation and design; the sensing infrastructure does the rest.