Track, identify, and analyze EV charging sessions in Home Assistant.
A custom integration for Home Assistant (installable via HACS) that turns your charger's raw sensor data into meaningful per-user charging records. Designed for multi-driver households sharing a single charger.
Key features:
- Plug-anchored session detection (go-e Gemini) — no phantom sessions
- RFID-based user identification
- Per-user energy and cost tracking
- Static and spot pricing support
- Guest charging with billing
- Session events for automations
- InfluxDB export for Grafana analysis
- Automatic RFID card discovery (go-e chargers)
- Session recovery after HA restart
Everything runs locally inside Home Assistant -- no cloud services required.
Before installing, make sure you have:
- An EV charger installed and connected to your local network
- Home Assistant 2025.3 or newer
- HACS installed
- A Home Assistant integration for your charger already set up (e.g. ha-goecharger-api2 for go-e chargers)
See docs/charger-profiles/goe-api2.md for detailed setup instructions, including recommended WebSocket mode configuration and entity mapping.
Any charger that exposes the following through a Home Assistant integration can be used:
- Car status sensor -- a sensor entity that reports when the vehicle is charging (e.g. "Charging", "Connected", etc.)
- Energy sensor -- session energy in Wh or kWh
- Power sensor -- current charging power in W
- RFID sensor (optional) -- reports the RFID card index or UID used to authorize a session
If your charger brand is not yet documented, see Charger Profiles (Community) for how to contribute a profile.
- Open Home Assistant and navigate to HACS > Integrations.
- Click the three-dot menu in the top right and select Custom repositories.
- Enter
cryptotomte/ev-charging-manageras the repository URL and select Integration as the category. - Click Add, then find EV Charging Manager in the HACS store and click Install.
- Restart Home Assistant.
- Go to Settings > Devices & Services > Add Integration.
- Search for EV Charging Manager.
- Follow the setup wizard:
| Step | What you configure |
|---|---|
| Charger profile | Select your charger brand (e.g. go-e) or "Generic" |
| Charger serial | A unique name/serial for this charger instance |
| Entity mapping | Map the charger's HA entities (car status, energy, power, RFID) |
| Pricing | Choose static or spot pricing and enter rates |
After completing the wizard, the integration creates a device and begins monitoring for charging sessions.
Users represent the people who charge at your location. Each user can have RFID cards mapped to them for automatic identification.
- Go to Settings > Devices & Services > Integrations.
- Find EV Charging Manager and click on your charger entry.
- Click Add User (subentry).
- Enter the user's name and type (resident or guest).
Guest users have additional billing options (custom price per kWh or markup factor).
Vehicles store battery specifications used for SoC estimation.
- Go to Settings > Devices & Services > Integrations.
- Find EV Charging Manager and click on your charger entry.
- Click Add Vehicle (subentry).
- Enter the vehicle name, battery capacity, usable capacity, and charging parameters.
RFID mappings connect a physical RFID card to a user (and optionally a vehicle).
Automatic discovery (go-e chargers): The integration reads the charger's card registry directly. When adding an RFID mapping, you select from discovered cards and assign them to a user and vehicle.
Manual fallback (other chargers): Enter the card index or UID manually based on your charger's documentation.
- Go to Settings > Devices & Services > Integrations.
- Find EV Charging Manager and click on your charger entry.
- Click Add RFID Mapping (subentry).
- Select or enter the card, then assign a user and optionally a vehicle.
| What | Who |
|---|---|
| Session detection and tracking | Integration (automatic) |
| RFID user identification | Integration (automatic) |
| Energy and cost calculation | Integration (automatic) |
| Session events for automations | Integration (automatic) |
| Per-user statistics | Integration (automatic) |
| Session recovery after restart | Integration (automatic) |
| Charger hardware installation | User |
| Charger network connectivity | User |
| HA integration for charger | User |
| User / vehicle / RFID setup in config | User |
| Pricing configuration | User |
| InfluxDB setup and maintenance | User |
| Grafana dashboards | User |
In short: you set up the hardware and tell the integration who your users are. The integration handles everything else.
The integration creates the following sensor entities per charger:
| Sensor | Description |
|---|---|
| Status | Engine sub-state: idle, waiting, charging, or charged |
| Current User | Name of the identified user (or "Unknown") |
| Current Vehicle | Name of the identified vehicle |
| Session Energy | Energy delivered in the current session (kWh) |
| Session Power | Current charging power (W) |
| Session Duration | Elapsed time of the current session (HH:MM:SS) |
| Session Cost | Accumulated cost of the current session |
| Estimated SoC Added | Estimated state-of-charge added to the battery (%) |
| Last Session Charging Duration | Active charging time in the most recent completed session (s) |
| Last Session Connection Duration | Total plug-in time in the most recent completed session (s) |
Session sensors become available when a charging session is active and return to unavailable when idle.
Last Session sensors become available after the first completed session.
For each configured user (and a combined "Guest" set), the integration creates:
| Sensor | Description |
|---|---|
| Total Energy | Cumulative energy charged (kWh) |
| Total Cost | Cumulative charging cost |
| Session Count | Number of completed sessions |
| Last Session | Timestamp of the most recent session |
| Average Energy | Average energy per session (kWh) |
Statistics sensors include a monthly_breakdown attribute with per-month totals.
| Sensor | Description |
|---|---|
| Charging | On when a session is actively tracking |
There are two complementary approaches for long-term analysis. You can use either or both.
Configure Home Assistant's built-in InfluxDB integration to log the EV Charging Manager sensor entities. This gives you continuous time-series data (power curves, cost accumulation, etc.) directly in InfluxDB.
This is entirely managed through HA's InfluxDB integration configuration -- no special setup in EV Charging Manager is needed.
For per-session summary records (one data point per completed charge), use the provided automation template:
- Copy
automations/ev_charging_manager_influxdb_export.yamlinto your Home Assistant automations. - The automation triggers on
ev_charging_manager_session_completedevents and writes a summary record to the InfluxDB measurementev_charging_sessions.
Each record includes: user, vehicle, energy (kWh), cost, duration, cost method, and timestamp.
See docs/grafana-queries.md for example Flux/InfluxQL queries to build dashboards.
When a session completes without a recognized user, it is attributed to "Unknown". Common causes:
- No RFID mapping -- the card used is not mapped to a user. Add an RFID Mapping subentry.
- trx sensor not updating -- the charger's RFID/transaction sensor is not reporting values. Verify the entity in Developer Tools > States.
- Card not active on charger -- some chargers require cards to be explicitly enabled. Check your charger's app or configuration.
The integration logs a diagnostic reason code for each unknown session. Check the Home Assistant logs for entries containing unknown_reason to pinpoint the cause.
The integration compares session-tracked energy against the charger's total energy counter (if configured). A warning indicates a deviation that usually means sensor updates were missed during the session (e.g. due to network issues or charger disconnects).
- Verify the charger's network connection is stable.
- If using polling mode, consider switching to WebSocket mode for more reliable updates.
The integration automatically recovers active sessions after a Home Assistant restart:
- The active session snapshot is persisted periodically (default: every 5 minutes).
- On startup, the engine restores the session if the same RFID card is still active.
- Very short sessions (< 60s) or sessions with negligible energy (< 50 Wh) may be discarded as micro-sessions.
Check the logs for "Recovered session" entries to confirm recovery occurred. If a session was lost, look for "Discarded micro-session" or "Session continuity mismatch" messages.
The plug-anchored session engine (go-e Gemini) tracks sessions using the cable presence sensor as the primary boundary signal. Key behaviors:
- Transient disconnects trigger a configurable grace timer rather than immediate session end.
- Charger power outages do not end sessions (all entities simultaneously unavailable).
- Multi-window sessions track charging duration separately from connection duration.
- Micro-sessions (< 60 s or < 50 Wh) are discarded automatically.
See docs/SESSION-MODEL.md for the full specification, including the session store schema and configurable timing parameters.
Charger profiles define how the integration maps a specific charger brand's entities to the internal data model. Profiles for supported chargers are documented in docs/charger-profiles/.
To contribute a profile for a new charger brand:
- Identify the HA integration and its entity naming conventions.
- Document the entity mappings (car status, energy, power, RFID) in a markdown file.
- Submit a pull request adding the profile to
docs/charger-profiles/.
See docs/charger-profiles/README.md for the template and guidelines.
This project is licensed under the MIT License.