android-telemetry-pipeline

What problem does it solve?

This Skill helps you design an Android telemetry system that reliably captures 10–100Hz sensor data without overwhelming SQLite, while still streaming live updates to the UI and preserving session structure for later analysis.

Core Features & Use Cases

• Hybrid storage design: Stores session metadata (Room) while writing raw telemetry samples to append-only CSV files on disk to avoid WAL/query churn.
• Session lifecycle management: Supports atomic session creation, clean finalization (end time + sample count), and post-session analysis status updates in Room.
• Telemetry normalization & streaming: Unifies BLE and phone sensor inputs into a single TelemetryFrame model, then uses StateFlow-backed UI state such as active source and recorded sample count.
• Buffered file writes: Flushes buffered samples periodically (e.g., every N frames) to reduce disk I/O per sample and prevent write contention.
• Post-session analysis pipeline: Parses the CSV after finalization on a background dispatcher and writes analysis results back to Room.

Quick Start

Use the android-telemetry-pipeline skill to implement a 25Hz BLE-to-UI telemetry recorder that creates a session, buffers and flushes samples to session_*.csv files, streams StateFlow updates to a Compose UI, and runs summary analysis when recording ends.