Intercity bookings never get their own train — they ride whichever import/export
train passes through their corridor — so the work is scattered across other
people's schedules and there was nowhere to see it as a whole. The per-schedule
ride-along panel answers "what can THIS train carry"; this answers "what is
happening to intercity cargo".
Purely additive: the existing ride-along panel and the schedule detail page are
untouched, and loading/unloading still happens there, where the train's position
is confirmed. This is a read-only view that points back to it.
Each row carries both ends' facility status, because a booking whose origin or
destination has no equipment can never be worked there — the operator should see
that while the train is still coming, not when the load is refused. Those
bookings are counted and called out.
New GET /train-scheduling/intercity/bookings; the type is IntercityRideAlongRow,
not IntercityBookingRow, which already means the per-schedule candidate row.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Which yards can handle cargo is data, not a constant — the set grows. The Yards
config screen now carries the flag, so adding a facility is a toggle rather than
a code change. Config-driven, so this is a column plus a form field.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Frontend for the per-truck EDR backend:
- Assign modal takes a container LOAD per truck (MultiSelect, max 2) instead of a
single container, so a truck can carry one 40ft or two 20ft. Sends
containerNumbers[]; the API enforces the size rule.
- requiredVehicles is now size-aware: a 40ft fills a truck (1 each), two 20ft
share one. It previously assumed ceil(containers / 2) regardless of size, so a
2x40ft booking under-counted the trucks needed.
- Bulk bookings show the drawdown in the assign modal — remaining vs total
tonnage, what's been hauled, and a "fully hauled" state when it reaches 0.
- New "Truck exit papers" row action opens a per-truck list (plate, containers,
arrival, exit, net weight) with a per-truck exit-paper download.
- last-mile findAll/findById now load each assignment's containers so the UI can
hydrate a truck's load.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- gateStats(): items cleared through the gate today, average arrival→gate
turnaround (hours, 30d), and clearances per hour over the last 24h → new
GateThroughputCard in the dashboard Performance section.
- Live board: every dashboard query (dashboard, ops, throughput, dwell, cycle,
on-time, zone occupancy, accrual, gate) now auto-refreshes on a 60s interval,
with a "Live" indicator in the header.
New endpoint GET /warehouse-inventory/gate-stats (guarded). Deferred (need
upstream data): capacity forecast, labour productivity, WebSocket push, yard map.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Adds the operational-performance layer to the warehouse cockpit:
- Dwell time & aging: dwellStats() (avg + 0-3/4-7/8-14/15+ buckets over
in-warehouse items) → DwellAgingCard histogram.
- Cycle time & on-time: cycleStats() (arrived→ready→loaded→dispatched stage
averages + dock-to-dispatch) and onTimeDispatchStats() (share of items that
left before their storage free-days expired, resolved via the fee engine's
own rule matching) → CycleTimeCard.
- Live tiles: opsStats() now returns receivedYesterday; KpiStrip renders an
optional ▲/▼ delta, and the ops strip shows received-today vs yesterday.
New endpoints: GET /warehouse-inventory/{dwell-stats,cycle-stats} and
/warehouse-fees/on-time-dispatch (all guarded). New "Performance" section on
WarehouseDashboardPage. On-time reads N/A when there is no sample / no active
storage rule.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
