Generator Monitoring System
Full-stack industrial monitoring system. Firmware collects 77 metrics from generators, transmits via GSM, visualized in real-time on a mobile app.
Programmed the generator chipset in C to collect and transmit 77 operational metrics (temperature, oil pressure, vibration, fuel, fault codes) via GSM every 2 seconds. Deployed across all clients of the company, each running multiple gensets independently.
Data hit a custom ASP.NET Core API on EC2 behind a reverse proxy, then got parsed and stored in a time-series PostgreSQL schema on RDS optimized for high-write throughput and fast historical queries.
Started with a native Android app in Java for monitoring, then rebuilt it in React Native when the system scaled and needed cross-platform support. Dashboards, push notifications for critical thresholds, and remote diagnostics.
The hard parts: GSM reliability under bad network conditions, tuning the database for burst writes from entire fleets transmitting simultaneously, and keeping the whole pipeline fault-tolerant with retry logic at the firmware level and failover policies on the backend.