Operations twin
Operational picture
Abstract support network / fictional hubs
AOC workflow
Illustrative five-division process model
Mission coordination queue 18
Support missions only / select a row for state details
| REQUEST / PURPOSE | ROUTE | STATE | DUE | ASSET |
|---|
Event stream
Observed model events
Test changes before applying them.
Fork the observed state into independent baseline and alternative branches. The live source is not changed by a forecast.
Alternative assumptions
Starts from the latest observed state
Deterministic estimates from illustrative rules. Not validated predictions, flight-planning guidance, or confidence intervals.
Backlog over the forecast horizon
Run a comparison to generate model trajectories
Two branches. One starting state.
Adjust the assumptions on the left and run the model.
Outcomes are computed, not pre-scripted animations.
Computed outcome comparison
Both branches start from the same frozen observation
| METRIC | BASELINE | ALTERNATIVE | CHANGE |
|---|
Model result, not an operational recommendation.
A real data boundary, not a hard-coded display.
The demo source and the observed twin are separate state objects. An external observation replaces the entire demo state; data sources are not silently mixed.
Source & synchronization
State delivery, provenance and freshness
Pausing telemetry freezes the twin while the simulated source keeps running. Resume delivery to catch up. Replay never changes the source.
Load a normalized observation
Import a validated JSON observation or a state previously exported by this application. Import opens a read-only snapshot, not a running real connection.
External observation adapter
Optional / read-only JSON polling
For the included local bridge, launch START_LOCAL_SERVER.bat and open its browser page first. The default standalone demo needs no server.
{
"schema_version": 1,
"source": "Your approved test adapter",
"observed_at": "ISO-8601 UTC timestamp",
"sequence": 1,
"state": { "modelVersion": "1.0.0", ... }
}Download the example for a complete, valid state. Invalid values, broken asset associations and out-of-order observations are rejected.
What this model actually computes
Transparent assumptions / replace with calibrated operational rules before real use
Workflow capacity
Requests move through five illustrative division queues, then wait for pooled aircraft. Priority and due time determine service order.
capacity = staff × 0.65 × availability factorsQueue work is reduced in one-minute steps. Division capacity depends on network, power and relevant application availability.
Resource & environment coupling
Aircraft cannot support overlapping missions. Flight and turnaround occupy a pooled asset. Nonlocal assets incur a fixed reposition allowance.
flight = (base + reposition) × (1 + weather / 200)Weather above 85/100, network below 25%, or power below 30% prevents new launches. These thresholds are invented for this demo.
Honest model boundaries
This is a synthetic digital-twin prototype, not a fielded AOC twin. The serial workflow, capacities, travel times and disruption rules are illustrative, not doctrine.
No targeting, weapons employment, flight deconfliction, classified feeds, security accreditation, calibrated forecasting or real command execution is implemented.