rtl_tracking

🔌 A high-performance Python utility designed to capture live JSON data streams from rtl_433, store them into an optimized SQLite database, and generate real-time tracking metrics.

Purpose

The primary objective of this project is to create a persistent storage layer for RF data packets intercepted via Software Defined Radio (SDR) using rtl_433.

This tool serves as an excellent practical framework for learning how to intercept standard output (stdout) data streams from compiled binaries and pipe them directly into automated Python processing loops.

Practical Use Case: TPMS & Sensor Tracking

By capturing data on the 433.92 MHz ISM band, you can permanently log and track:

• TPMS (Tire Pressure Monitoring Systems): Monitor and identify unique sensor IDs, tire pressures, and temperatures from vehicles driving around your receiver.
• Weather Stations & Smart Meters: Aggregate historical temperature, humidity, and power data from residential remote sensors.

Key Features

• Stream-Optimized Ingestion: Uses a single, persistent SQLite connection channel instead of spawning an I/O request per line. This completely eliminates data loss and packet drops during intense RF bursts.
• Embedded Storage Optimization: Automatically applies SQLite Write-Ahead Logging (PRAGMA journal_mode = WAL;) and relaxed synchronous boundaries. This drastically improves write speeds and extends the physical lifespan of flash storage (MicroSD cards on a Raspberry Pi Zero 2).
• Automated Analytics View: Deploys a relational database view called tracking that automatically groups, counts, and sorts unique device identifiers over time.

Requirements

You must have an SDR device (such as an RTL-SDR Blog V4) and the rtl_433 binary installed on your system.

Usage and Routing Options

Live SDR Interception

Pipe the live JSON outputs from rtl_433 directly into the tracking pipeline:

rtl_433 -F json | python3 rtl_tracking.py

Logging Outputs with Background Persistence

Store the tracking terminal feedback into a local log file:

rtl_433 -F json | python3 rtl_tracking.py >> output.log

Batch Processing Historical Captures

You can back-populate your database by passing any pre-recorded newline-delimited JSON log file:

cat my_json_capture.json | python3 rtl_tracking.py

Filtering Specific Devices

To optimize CPU cycles on lower-powered devices (like a Raspberry Pi), it is highly recommended to filter your rtl_433 capture targets directly at the source using the -R argument (device protocol IDs):

# Example: Listen exclusively to standard TPMS protocols (e.g., Schrader, Renault, Toyata)
rtl_433 -R 60 -R 110 -R 195 -F json | python3 rtl_tracking.py

Review the complete protocol registry list directly on the rtl_433 repository.

Database Architecture

• data (Table): Logs raw incoming telemetry coordinates including reception timestamp (time), protocol description (model), and the unique sensor identifier (id).
• tracking (View): A virtual evaluation layer compiling operational metrics. It summarizes total detection counts per hardware sensor ID, sorted dynamically by protocol models.

Extending the Project

Because the data mapper utilizes a dynamic keys allocation dictionary wrapper (data.keys()), you can easily adapt this project to log additional fields (such as pressure_PSI, temperature_C, or battery_ok) by simply appending those fields to the initial prepare_database schema declaration dictionary.