The most valuable view of a charging network is the one you can read in a glance: where is every plug right now, which sessions are live, and what is actually broken. Until now, that view only existed scattered across per-station pages and server log lines.

ChargePeer now ships a Live Charger Status dashboard for charge-point operators. Open /provider/dashboard and every station you own renders as a single tile, color-coded by derived state: available, charging, fault, offline, or busy. Each tile carries its connector inventory (plug types and power), the operator-set rate, the location, and when the station was last seen.

The status shown is not a guess — it is derived at request time from real state. A station is marked fault when it has an open hardware alarm on record. It is offline when it has stopped heartbeating (or the operator has turned it off). It is charging while a booking is actively delivering energy, and in that case the tile shows the live session: the driver, requested kWh, accrued cost, and progress bar. Everything else resolves to available or busy based on the operator's own availability flag.

Hardware alarms were the biggest blind spot. When a charger reported an alarm before, the backend logged an error line and moved on — nothing was persisted, so nothing was ever surfaced to the operator. Alarms are now recorded as open faults, and the dashboard lists each one with its type, severity, and report time. When the hardware is fixed, the operator marks the fault resolved in one tap and the station automatically returns to the network once no other faults remain.

The dashboard refreshes automatically every 15 seconds, so operators watching a busy lot see it evolve in real time without touching anything. The summary bar up top totals the network: available, charging, faults, offline, and busy counts at a glance.

Two quieter fixes came along with it. The provider-role check now matches roles case-insensitively and always admits Super Admin accounts, so no operator is ever locked out of their own network. And the vehicles schema migration now scopes its column checks to the active schema — a latent bug that could silently leave the live-status columns missing on the authoritative table is gone, which also un-breaks the existing availability toggle for operators whose accounts were affected.

This is the foundation for more network tooling: fleet-level power draw, per-station utilization, and predictive fault alerts are all natural next steps now that the live state model exists.