Skip to content
pubscout

offline_sos_system

0.0.6, released 5 weeks ago by bhagyaprasad.dev

Detect crashes and anomalies on‑device using accelerometer/gyroscope data and a TFLite model.

$ flutter pub add offline_sos_system
Source repository

What it does

The package continuously reads sensor telemetry, runs an on‑device TensorFlow Lite model, and emits a CrashEvent stream when a crash is inferred. It also provides configurable thresholds and cooldown handling.

AI-written summary of version 0.0.6, based on the package's README and metadata. Report an inaccuracy

Use it when

  • You need zero‑latency crash detection that works without network connectivity
  • Your app must run entirely offline for safety‑critical scenarios
  • You want a ready‑made headless engine to plug into your own UI or emergency workflow

Skip it when

  • You rely on cloud‑based analytics or remote model updates
  • Precise crash classification requires more sensors than accelerometer/gyroscope
  • You need a full safety system with UI components built‑in

Strengths

  • 100 % offline inference with no network dependency
  • Highly configurable thresholds and cooldown logic
  • Simple three‑line integration and stream API

Weaknesses

  • Limited to accelerometer and gyroscope data; may produce false positives on varied hardware
  • No built‑in UI or emergency action handling – you must implement it

Setup

Kind
Flutter widgets
Effort
Low
Code generation
No
Direct dependencies
2

Add sensors_plus and tflite_flutter dependencies (already included). Request sensor permissions on Android/iOS if required by the platform.

import 'package:offline_sos_system/offline_sos_system.dart';

final engine = SafePulseEngine();
engine.onCrashDetected.listen((e){print('Crash at ${e.timestamp}, prob ${e.probability}');});
await engine.start();

Alternatives

Packages that do the same job and could replace offline_sos_system.

  • face_detection_tflite

    Advanced face & landmark detection, embedding and segmentation using on-device LiteRT (formerly TensorFlow Lite) models.

  • pose_detection

    Pose, person and landmark detection using on-device LiteRT (formerly TensorFlow Lite) models.

  • hand_detection

    On‑device hand detection, landmarks and gesture recognition using LiteRT models.

Works with

Packages commonly used together with offline_sos_system: ones it builds on, ones built on it, and its own family.