Reading a supertall
as a set of thresholds

The section is the argument: operational thresholds cluster by program. Node counts per band are indicative. The core — lifts, stairs, risers — runs the full height and carries the vertical-transport and MEP nodes. Amber = Type 2 (safety/auth), blue = Type 1.

Where the public, four operators, and back-of-house meet. Nodes sit at each arrival threshold and each goods/waste threshold — never on people, always on the point where an operation is confirmed. Amber node = Type 2 (safety).

A Ritz-Carlton residential level (L23–73). Units ring the core; the pixelated void becomes a cantilevered terrace. Nodes sit at the service spine and the terrace threshold — the facade cut is itself a maintenance and safety point — an amber Type-2 node.

A Standard guest level (L1–18). Double-loaded corridor; the housekeeping and service spine is where "room ready", "floor cleaned" and "linen restocked" become witnessed records rather than assumptions. Amber = Type 2 (vertical transport).

Each node is a passive NFC tag on a surface. A phone tap is the only data channel. The record is hash-chained and tamper-evident. Every task is events, not people — a role confirms a threshold; no individual is tracked.

Most of the tower is the standard passive node. The safety- and authenticity-critical points get the active chip; the few places where a live value matters get the sensor chip. Counts are an illustrative full-stack estimate; an MVP starts with a representative subset.

A node is a passive sticker on a surface — on in minutes. No wiring, no cameras, no surveillance. The human tap is the only sensor. The building stays sovereign: the record is the owner's, held by no one else.

A chip is passive and small; the socket is the housing that seats it flush into a surface, protects it, and makes the tap durable and legible. Sockets are made in the DWNTWN atelier — the leather line doubles as the socket line.

Each node opens its own one-question screen on a tap. No app, no login. Active-chip screens (Type 2) confirm with the chip's own LED + buzzer.

Every tap appends one tamper-evident record. Over a year the operational memory compounds while documented risk drops — the asset gets a provable history it never had.

How often a node is actually touched — indicative, per shift and per day. Enough to be useful; light enough to disappear into the work.

Illustrative figures — a live deployment sets its own baseline over the first weeks.

The operator's view. Every witnessed tap compounds; after a year the building holds a record it never had — and the record starts to explain itself.

The constitution the system is bound by — events, not people. Ten unamendable prohibitions, published to the public domain. What the building may never become.

Kind feedback from the trenches — architects, FM, engineers, insurers, operators who have actually run buildings like this. Answer what you can, add a line of context to any question, and submit. It comes straight to the team. No login.

Whether your building is on the drawing board or twenty years in service, a small pilot fits: a few nodes, one zone, 90 days — witnessed operational memory you own. A few details and we'll be in touch.