depiprojectdata/DEPI-Project

End-to-end real-time data engineering platform monitoring 3,000 IoT sensors across Egypt's 27 governorates. Detects gas leaks, fire risk, and air pollution using Kafka streaming, Spark processing, and SQL Server storage. Visualized through Grafana and a custom React dashboard. # 🏙️ Smart City Gas & Air Safety Monitoring Platform ### Real-time IoT data engineering pipeline protecting 3,000 buildings across all 27 governorates of Egypt *Graduation Project — Digital Egypt Pioneers Initiative (DEPI) 2026* Overview • Architecture • Data Flow • Components • Getting Started • API ## 📖 Overview ### The Problem Gas leaks, LPG explosions, and unmonitored air quality are a silent, everyday risk in dense residential and industrial areas. In a city the scale of a nation like Egypt — millions of homes, mixed building ages, aging gas infrastructure, and industrial zones sitting next to housing — three failure modes repeat constantly: - **Gas & LPG leaks** go undetected until concentration is already inside the flammability range. - **Fire precursors** (rising heat + smoke) are noticed by people, not systems, which means minutes are lost. - **Air quality degradation** (CO, CO₂, PM2.5/PM10, AQI) is chronic and invisible — there is no live, governorate-level visibility for operators or citizens. Civil protection and utility operators today mostly react *after* an incident is reported by a person. There is no unified, real-time, city-scale telemetry layer that watches every building continuously and tells operators **where**, **why**, and **how urgently** to respond. ### The Solution This project is an **end-to-end real-time data engineering platform** that simulates a nationwide IoT sensor deployment — **3,000 sensor nodes across Egypt's 27 governorates** — streaming methane, LPG, smoke, CO, AQI, CO₂, PM2.5/PM10 and temperature readings every 30 seconds, running them through a proper **Bronze → Silver → Gold (Medallion) pipeline**, and serving the results through two live dashboards and a bilingual AI safety assistant. It is built to demonstrate a production-shaped data engineering pattern, not a toy demo: | Capability | How it's delivered | |---|---| | Ingest at scale, in order | Kafka topic partitioned by `house_id` | | …