# Hub operations diagram — version 4.2

Run the model and open **2 Simulation**. This view shows a schematic corridor, hub entry, queue staging, charging bays, driver-rest building, outbound road, power connection and transformer. It is an interactive engineering diagram, not a geographic map or an installation design. Bay positions, road lanes and driver amenities are illustrative. All vehicles/bays are counted even when only some are drawn.

## Main scenario comparison

Use the Scenario selector to compare priority, occupancy and vehicle settings at the same model time. Play or scrub the clock. Change the look-ahead window to show upcoming arrivals. The clock-start field changes displayed time-of-day only; it does not alter arrivals or dispatch. The data remain minutes since study start. Road placement is inferred from arrival ordering for illustration: the comparison input does not contain GPS or road distances. Approach battery bars in this mode show the supplied arrival-SOC assumption, not a measured driving SOC.

Approaching trucks enter the diagram, wait in staging, occupy a charging bay and depart when the engine releases it. Green means charging, purple means energy-complete but bay held. Actual early bay release under an idle-yield policy is shown. Firm break time is retained. Queues are not physically capacity-limited by the number of spaces in the drawing. Illustrative bay assignments are not charger bookings or real bay IDs.

Select a truck by clicking its icon or using **Inspect truck**. Its timeline shows arrival, charging start, energy completion and bay release. Jump buttons move the simulation to its arrival or charging start. The dispatch board lists active, waiting and nearby future trucks; filter IDs to find one quickly. The event board shows the next forecast arrivals, charging starts, completions and releases in the look-ahead window. Unknown times remain explicitly unknown at the study cutoff. A nominal release after the cutoff is a planning estimate subject to the configured rules and excludes future unknown arrivals.

The cross-scenario table compares queue length, charging/held/free bays, current demand and next release at the same time. A lower wait metric does not imply every truck is better served; check the backlog and energy delivered in Results.

## Approach and live observations

In **5 Approaches & live**, compute/import observations and choose **View on operations diagram**. The Data source selector switches between that report and the main comparison. Select a hub when the report includes multiple hubs. Simulation reports can be played through their known-arrival forecast. The clock is independent of the comparison clock. Road distances come from the supplied approach inputs; the diagram is still not geographically scaled.

Live reports show an assessed observation snapshot and its forecast. Live playback/scrubbing is disabled: the website does not invent vehicle movement between feed updates. The existing connector can refresh the view from new observations while open. Source labels and assessment time are shown, and stale data are flagged. If truck or hub data expire, refresh the observations rather than interpreting the old view as a real-time reading. Mock observations remain visibly labelled. Disconnecting a feed does not make its last report current indefinitely.

Live states between observations are model forecasts; arrivals require confirmation in updated hub snapshots. No GPS, routing service or real charger feed has been configured by default. See LIVE_DATA_GUIDE.md for the feed contract. The operation report now includes time-aligned per-truck charging/release forecasts and grid event intervals. Live report times are relative to the assessment; absolute ISO release times are also exported. The normalized hub rows use SOC measured at snapshot time for connected jobs.

## Grid and electrical indicators

The feeder/transformer panel shows the configured site grid limit, computed grid-side draw, battery-side charging, conversion loss and spare headroom. Power sharing follows the same vehicle caps, charger limits and efficiency as the Python simulation. A charged holding vehicle draws zero charging power while retaining its bay. No building loads, auxiliary site demand, generation or storage are invented.

The power chart distinguishes computed history (solid), future scenario demand (dashed), other scenarios (faint) and the selected site's supply limit. Energy is integrated from the computed grid intervals. Observation energy covers the supplied snapshot window; comparison energy starts at study start. Beyond the calculated horizon, uncomputed load is unknown rather than asserted to be zero.

These are load calculations, not a load-flow network model. Voltage, upstream capacity, faults, thermal overloads and power quality still require the power-system study. A configured limit is an assumption, not a verified grid connection agreement.

## Usability and validation

Use Map zoom or Expand view for a larger diagram. Small screens can scroll the map horizontally; all forecasts also appear in accessible text/tables. Motion respects reduced-motion preferences. Truck ID selection provides an alternative to clicking the canvas.

The JavaScript view is checked against 44 native-engine snapshots for queue/charging/holding states, occupied bays, grid and battery power, SOC bounds, event ordering and cutoff handling. Python tests cover operation report alignment, release forecasts and stale hub suppression. Browser checks cover scenario switching, observation mode, zoom, timelines and console errors. This validates implementation consistency; it does not calibrate transport or driver behaviour against real-world operations.