Jet1090 Plugin¶

The tangram_jet1090 plugin is the primary tool for integrating Mode S and ADS-B data into the tangram framework. It processes raw aviation surveillance data from a jet1090 instance and makes it available for both real-time visualization and historical analysis.

Overview¶

This plugin consists of:

A background service (planes) to maintain a real-time state of all visible aircraft and persist their data for historical queries.
A REST API endpoint (/data/{icao24}) to fetch the full, time-ordered trajectory for a specific aircraft.
A frontend widget to display a placeholder in the UI.

`planes` Service¶

The planes service is the core of the real-time functionality. It is a Rust-based component, wrapped with PyO3 for integration into the Python ecosystem.

sequenceDiagram
    participant J as jet1090
    participant R as Redis
    participant P as planes service
    participant C as channel service
    participant F as Frontend

    J->>R: PUBLISH jet1090 (raw message)
    P->>R: SUBSCRIBE jet1090
    Note over P: Process message, update state vector
    P->>R: XADD history:ingest:jet1090, *, ...

    F->>C: PUSH system:bound-box
    C->>R: PUBLISH from:system:bound-box
    P->>R: SUBSCRIBE from:system:bound-box
    Note over P: Update client's visible area

    loop Every second
        Note over P: Filter aircraft by each client's bbox
        P->>R: PUBLISH to:streaming-client1:new-jet1090-data
    end

    C->>R: SUBSCRIBE to:streaming-client1:*
    C->>F: PUSH new-jet1090-data

Continuous Tracking: It subscribes to the jet1090 Redis channel to receive decoded aircraft messages.
State Vector Maintenance: It maintains a comprehensive in-memory view of the current air traffic situation by collecting and processing state vectors for all active aircraft.
History Persistence: It acts as a producer for the tangram_history plugin. It batches raw messages into arrow Recordbatches and sends them to a redis stream. The history service consumes this stream and persists the data into a delta lake table.
Client-Specific Filtering: The service listens for bounding box updates from each connected frontend client. It filters the aircraft data for each client, sending only the aircraft visible within their map view.
Data Publishing: Once per second, it publishes the filtered state vectors to a dedicated Redis channel for each client (e.g., to:streaming-<client_id>:new-jet1090-data), which are then relayed to the browser via the WebSocket channel service.

Redis Events¶

Direction	Channel	Event/Command	Payload
Input	`jet1090`	`PUBLISH`	Raw JSON message from `jet1090`.
Output	`to:streaming-{id}:new-jet1090-data`	`PUBLISH`	`{ "count": 123, "aircraft": [...] }` containing visible aircraft.
Output	`history:ingest:jet1090`	`XADD`	Apache Arrow record batch (binary).

Trajectory API¶

The plugin provides an API for querying the historical data persisted by the planes service.

Endpoint: GET /jet1090/data/{icao24}

Retrieves all historical data points for the given aircraft icao24 by querying the delta lake table managed by the history service.

This endpoint is used by the frontend to draw historical flight paths and populate data charts when an aircraft is selected.

Configuration¶

To use this plugin, you must have a running jet1090 instance publishing data to Redis. You can also configure the plugin in your tangram.toml::

[core]
plugins = ["tangram_jet1090", "tangram_history"]

[plugins.tangram_jet1090]
# the redis channel that jet1090 is publishing to.
jet1090_channel = "jet1090"
# how long (in seconds) to keep an aircraft in the state vector table
# after its last message.
state_vector_expire = 20

# history persistence settings (requires `tangram_history` plugin)
history_table_name = "jet1090"
history_flush_interval_secs = 5
history_buffer_size = 100000

See [tangram_jet1090.PlanesConfig] for more information.