"AquaIntel: Multi-Sensor IoT Telemetry for Smart Water Management & Anomaly Detection"

"This dataset provides 3,886 multi-sensor time-series readings collected from a custom IoT Edge Node (Arduino Uno) engineered for smart water management and agricultural anomaly detection. Telemetry was recorded across three distinct physical soil profiles: Grey Sand, Red Sand, and Normal Soil. The physical hardware array integrates an analog resistive water level sensor, a YF-S201 inline flow rate sensor, a YL-69 soil moisture probe, and a DS18B20 water temperature sensor. Data was streamed continuously to a Python\/Flask edge gateway.Unlike purely simulated environments, this dataset captures highly valuable, authentic hardware behaviors and faults. Key features include explicitly labeled normal irrigation cycles, the successful execution of an automated 3-second hardware pump failsafe to prevent flooding, a real-world sensor displacement fault where the water level probe temporarily lost contact, and a runaway pump event resulting in total soil saturation. Ground truth labels (0_Baseline and 1_Normal_Irrigation) are provided for every timestamp. This dataset is designed for researchers building and benchmarking supervised machine learning classification models (e.g., Random Forest, Decision Trees) for precision agriculture, leak detection, and edge-level fault identification."