PUBLIC EVIDENCE COMMONS / EXPERIMENTAL

How much pressure is building at Terminal 3?

One transparent estimate from flights, checkpoints, delays, weather, and a privacy-filtered local evidence pipeline. It is a directional signal—not a headcount, queue promise, or airport instruction.

NOW / T3 + H

WAITING
—/100

LOADING

Reading the current public airport signals.

MODELflight-presence-kernel-1.0

TODAY / H-CONCOURSE FLIGHT PULSE

See the waves before they arrive.

Quiet Steady Busy Very busy

Building today's H-concourse flight pulse…

Selected time—
Flight pulse—
Relative load—

Departures peak about 75 minutes before gate time; arrivals create a smaller, shorter pulse.

THE READING, OPENED UP

Four signals. No mystery score.

Missing feeds lower coverage instead of being silently replaced with guesses. Each signal keeps its own meaning.

01
—

H flight wave

Scheduled and updated H-gate arrivals and departures, translated into a concourse-presence window.

55% configured weight
02
—

T3 checkpoint

The highest current public wait reading across Terminal 3 checkpoints.

25% configured weight
03
—

ORD disruption

Arrival and departure delay pressure that can increase dwell and irregular flow.

15% configured weight
04
—

Flight weather

KORD flight category, precipitation, and wind—not a passenger count.

5% configured weight

LOCAL PIPELINE / PUBLIC ALLOWLIST

Use more evidence without publishing the private world.

UAI compiles a narrow public packet from local receipts. The site gets coverage counts, model states, and verification gates. It never gets raw rows, identities, staffing values, or H9 transaction and sales values.

PUBLIC PACKET97461db4709b…compiled 877 hr ago
HISTORY GATESTILL LEARNING

Promotion waits for both coverage thresholds and held-out performance against the current flight baseline.

Local dates8 / 28
155 public-source snapshots retained as aggregates
Distinct date-hours162 / 448
2026-08-22 through 2026-08-30
PIPELINE RECEIPTS2/8 readydiagnostic projection incomplete
  • Interval projectiondegraded
  • Transaction-demand modeldegraded
  • Flight-opportunity basisdegraded
  • Private sales-validation bridgedegraded
  • Direct aggregate pressure labelsdegraded
  • Aggregate observation planready
  • Historical T100 route-mass priorready
  • Delayed aggregate settlementdegraded
PROSPECTIVE EVALUATIONcontacting ledgerawaiting durable readback
Frozen claims—— awaiting labels
Settled— / 56— / 14 dates
Flight baseline MAE—candidate —
Correction latency—— retained corrections

Each claim is frozen 30–60 minutes ahead. Scores remain blank until a delayed aggregate label settles the same target; no self-score or automatic promotion is allowed.

AIRPORT ADAPTER / 001ORD

Terminal 3 / Concourse H

flight schedule and status
live
checkpoint waits
live when available
airport operations
live when available
aviation weather
live
local public history
diagnostic
direct aggregate labels
shadow
PORTABILITY CONTRACT1 live reference · 1 source-verified shadow

Collect MDW terminal-scope history and direct aggregate labels before any pressure calibration.

MDWContext live · pressure abstained ↗

H CONCOURSE

17 official gates
—

gate events in the next 2 hours

Departures
—
Arrivals
—
H9 events
—
Remaining-day peak
—

An event is an H-gate arrival or departure—not one passenger and not one sale.

TERMINAL 3 CHECKPOINTS

live public readings
—

highest current wait

No usable T3 checkpoint readings right now.

Waits can change quickly and checkpoint hours vary. Follow airport and TSA instructions.

ORD OPERATIONS

FAA status checked
—

larger average delay

Arrivals
—
Departures
—
Cancellations
—

Delays can increase dwell while cancellations can reduce later passenger volume.

AIRFIELD CONTEXT

not scored
—— wind
— mi visibility
Aircraft within 25 mi
—
Reporting on ground
—

Open aircraft coverage is useful context but does not identify a terminal, gate, or passenger count.

H9 / SALES VALIDATION LANE

Tie pressure to sales without publishing the business.

The public pressure series and H9 demand must stay separate until a privacy-safe aggregate proves whether one helps predict the other.

PRIVATE VALIDATION BRIDGEdegraded

Its status may be public. Its dollars, orders, customers, employees, and thin buckets are not.

01Pressure bucketPublic 30-minute index
02Lead / lag testDoes pressure lead demand?
03H9 demand indexMedian comparable bucket = 100

Release only minimum-count-suppressed, normalized buckets. Show error and sample size beside any claimed relationship; a click or a single shift is not validation.

Download safe aggregate schema ↓

GORPTASTIC / CIVIL EPISTEMIC BUS

The packet can travel. Its authority cannot grow in transit.

Every reusable signal says what was observed, what was inferred, what would falsify it, and how it can be corrected. Other worlds may cite the evidence; they cannot silently turn it into permission or action.

01 / OBSERVATION0 public feeds reachedAggregated airport facts
02 / INFERENCEAwaiting live readingRelative pressure, not people
03 / HYPOTHESISDoes pressure lead H9 demand?Unpromoted and tested privately
04 / CORRECTIONSupersede, retain, replayNo silent history rewrite
H9

Validation lane

Compare lagged, suppressed aggregate demand with pressure. Publish performance, not the business.

HYPOTHESIS · NO CAUSAL CLAIM
ANY AIRPORT

Portable adapter

Reuse the contract only after local sources, terminal scope, freshness, and calibration pass.

METHOD CANDIDATE · NOT UNIVERSAL YET
TRADING

Context, never inheritance

A trading world may cite verified aviation context, but must form and test its own market hypothesis.

NOT A SIGNAL · NO ORDER AUTHORITY

METHOD / V2

What this can—and cannot—tell you.

ORD Pressure estimates relative passenger-facing pressure. It does not observe people directly, infer identities, or claim that airport traffic equals H9 sales.

MEASURED

Public observations

  • H-gate flight times and status
  • Terminal 3 checkpoint waits
  • ORD delays and FAA programs
  • KORD weather and open aircraft context
  • Aggregate local history coverage and receipt state
MODELED

Relative pressure

  • Departure presence peaks 75 minutes before gate time
  • Arrivals create a smaller post-arrival pulse
  • Available components are visibly weighted
  • Missing data reduces source coverage
  • Candidate models must beat the simple baseline
NOT CLAIMED

Things we cannot see

  • Terminal or concourse headcount
  • Queue length, store wait, or staffing capacity
  • Passenger destination inside Terminal 3
  • Current H9 sales or conversion
  • Aviation-to-market causation or trade authority

SOURCE LEDGER

0/6 feeds reached

Updated when the first live reading completes.