offline_sos_system 0.0.1

A pure Dart, offline-first Flutter package for on-device crash and anomaly detection using hardware telemetry and TFLite.

offline_sos_system

A pure Dart, offline-first Flutter package for on-device crash and anomaly detection.

This package orchestrates device hardware sensors (accelerometer, gyroscope) and TensorFlow Lite (tflite_flutter) to detect physical crashes in real-time. It operates completely offline, ensuring maximum privacy and zero latency.

Note: This is a headless detection engine. It explicitly does not handle UI, GPS location, or SMS dispatching. It simply analyzes telemetry and emits a stream of CrashEvent objects, leaving the application layer in full control of the business logic.

Features

• Offline-First: Zero network requests. All ML inference happens on-device.
• Bring Your Own Model: Bundles a default crash_model.tflite, but allows injecting custom trained models.
• Decoupled Architecture: Exposes a simple Stream<CrashEvent> without forcing native permissions (like SMS or Location) onto your app.

Installation

Add this to your package's pubspec.yaml file:

dependencies:
  offline_sos_system: ^0.0.1