A High Fidelity Synthetic IoT Dataset for Smart Waste Management (10 Million Records)

This dataset provides 10 million simulated records of smart waste bin levels across major neighborhoods in Dhaka, Bangladesh (Gulshan, Motijheel, Dhanmondi, Mirpur, and Old Dhaka). Unlike generic datasets, this simulation incorporates regional socio-cultural factors, such as the "Friday Mosque Effect" (spikes in waste generation during prayer times), weekend patterns (Friday to Saturday), and extreme seasonal events like Eid-ul-Adha. The data also includes real-world IoT challenges, such as sensor noise, battery depletion patterns influenced by weather conditions, and missing data points resulting from connectivity failures. It is designed for researchers working on time-series forecasting, anomaly detection, and smart city logistics.

Key Features:
1. 10 Million Records: High-volume time-series data.
2. Spatial Integrity: 5,000 unique bin IDs locked to specific Dhaka coordinates.
3. Socio-Cultural Seasonality: Includes the "Friday Mosque Effect" and "Eid-ul-Adha" seasonal spikes.
4. IoT Realism: Simulated sensor noise, battery drain, and connectivity failures.
5. Target Variables: Dual targets for ML Classification (needs_collection) and Regression (overflow_kg).