S-Bahn München Live Home Assistant Integration

Home Assistant integration for S-Bahn München public transport departures, covering all 152 known S-Bahn stations.

Features

• Real-time departure information
• Config flow with a searchable station dropdown — no more guessing station IDs
• Filter by destination and line
• Delay information
• Platform information
• Multiple stations, each as its own config entry
• Shared polling per station via a DataUpdateCoordinator

Installation

HACS (Recommended)

Click the button above, then "Add" to add this repository to HACS. Alternatively, add it manually:

1. Open HACS in your Home Assistant instance
2. Go to "Integrations"
3. Click the three dots in the top right corner
4. Select "Custom repositories"
5. Add this repository URL
6. Category: Integration
7. Click "Add"

Either way, afterwards:

8. Search for "S-Bahn München Live" in HACS and install it
9. Restart Home Assistant

Manual Installation

1. Copy the custom_components/sbahn_muenchen_live directory to your Home Assistant config/custom_components directory
2. Restart Home Assistant

Verify

• Go to Developer Tools → States
• Look for the new sensor.* entity for the station you added
• The state should show minutes until next departure

Configuration

Go to Settings → Devices & Services → Add Integration and search for "S-Bahn München Live". Pick a station from the dropdown and, optionally, a custom name for the sensor.

After setup, use the integration's Configure button to adjust:

Option	Default	Description
destinations	All	Comma-separated list of destinations to filter
lines	All	Comma-separated list of lines to filter (e.g. S1, S8)
timeoffset	0	Minimum minutes until departure
number	5	Maximum number of departures to show

Note: destinations matches as a case-insensitive substring, not an exact string — this is intentional: the same physical destination is reported with different text depending on which station asks. For example, a train to the airport shows up as "Flughafen/Airport ✈" at München Hbf, but as "Freising, Flughafen ✈" or "Flughafen ✈, Flughafen ✈" at München Ost. A filter of "Flughafen" matches all of these; the full exact string would silently miss most of them depending on the station. lines still matches exactly (e.g. S1), since line names don't vary like this. If in doubt, set up the sensor without filters first and check its departures attribute in Developer Tools → States to see the real strings.

Legacy YAML configuration

Existing configuration.yaml entries under sensor: - platform: sbahn_muenchen_live are automatically imported as config entries on startup (with a repair notification asking you to remove the YAML block afterwards). New YAML configuration is not needed — use the UI going forward.

sensor:
  - platform: sbahn_muenchen_live
    nextdeparture:
      - station: "Flughafen Besucherpark"
        name: "Airport Departures"

Sensor Attributes

The sensor provides the following state and attributes:

State: Minutes until next departure

Attributes:

• destination: Destination of the next departure
• line: Line name (e.g., "S1", "S8")
• type: Transport type (e.g., "S-Bahn")
• platform: Platform number
• delay: Delay in minutes
• icon: Material Design Icon
• train_units: Number of coupled S-Bahn units (1–3) for this train, if known (see note below)
• departures: List of all upcoming departures
• messages: Service messages (currently empty — the geOps timetable endpoint doesn't provide any)

Note on train_units: geOps only reports train formation (how many units are coupled together) for vehicles currently broadcasting live position data, via a separate realtime feed the integration keeps a single shared background connection open for (covering the whole network, reused across all configured stations). For trains without an active live position yet — usually further in the future — train_units is null until that data arrives, closer to departure time.

Example Lovelace Card

type: entities
title: Munich Airport S-Bahn
entities:
  - entity: sensor.airport_departures
    secondary_info: last-updated

Advanced Lovelace Card with Departures List

type: custom:auto-entities
card:
  type: entities
  title: Airport Departures
filter:
  template: |
    {% for departure in state_attr('sensor.airport_departures', 'departures') %}
      {{
        {
          'entity': 'sensor.airport_departures',
          'name': departure.line + ' → ' + departure.destination,
          'secondary_info': departure.time_in_mins | string + ' min' + (' +' + departure.delay | string + ' min' if departure.delay > 0 else '')
        }
      }},
    {% endfor %}

Troubleshooting

No Departures Showing

1. Check that the selected station has current S-Bahn traffic
2. Verify your internet connection
3. Check Home Assistant logs for errors: Settings > System > Logs
4. Try increasing the number option to fetch more departures

Integration Not Loading

1. Ensure all dependencies are installed (websockets, certifi)
2. Check Python version (requires Python 3.11+)
3. Restart Home Assistant after installation
4. Check for errors in the logs

Sensor State is "Unknown"

1. Check if the API is reachable: wss://api.geops.io/realtime-ws/v1/
2. Verify there are departures at the current time for that station

Data Source

This integration uses data from the S-Bahn München Live Map provided by geOps.

Why this integration?

The timetable data from the geOps Live Map API is more accurate and reliable than other public transport APIs because:

• Real-time updates: Direct connection to the live tracking system
• Precise timing: Actual timetable data with delay information
• High reliability: Used by the official S-Bahn München live map
• WebSocket connection: Instant updates without polling delays

The integration specifically uses only the timetable endpoint (GET timetable_{station_id}) for optimal performance and accuracy, rather than map/trajectory data. The station list (custom_components/sbahn_muenchen_live/stations.py) was generated once from the GET station_schematic endpoint, since geOps has no by-name station search API.

License

MIT License

Credits

• Data: geOps.io
• Integration structure inspired by the MVG integration

Support

For issues, feature requests, or questions, please open an issue on GitHub.