Design and Validation of an IoT-Enabled Safety Monitoring System for Agricultural UAV Operations
The challenges faced by Unmanned Aerial Vehicles (UAVs) used in precision agriculture are growing. Operational risks arise from harsh conditions, such as abnormal flight behaviors, extended missions, and environmental circumstances. Current flight controllers primarily focus on navigation and stabilization, with only limited ability to monitor real-time safety. This study presents a low-cost, real-time safety monitoring platform for agricultural UAVs using an ESP32 microcontroller integrated with an MPU6050 inertial measurement unit (IMU) and a DHT11 temperature sensor. The system continuously measures roll, pitch, and on-board temperature, sending telemetry data to a cloud-based IoT platform for visualization and alerts. An empirically derived, threshold-based anomaly detection plan, based on over 100 normal flight data points, more than 20 Pixhawk crash log analyses, and observations of climatic conditions from 28 Indian states, allows for timely detection of abnormal operating conditions. The monitoring module has a sampling rate of 30-40 Hz and uses persistence-based detection logic to ensure that sustained anomalies are present over a 125-165 ms timespan. It was experimentally verified that roll and pitch estimates based on an atan2-based approach have a mean absolute error of 7.62° and 13.52°, respectively, compared to reference flight controller measurements. Temperature validation shows that the DHT11 sensor's temperature readings do not exceed the ±4.6°C deviation limit relative to the reference readings. These findings reveal that the suggested framework offers a scalable, cost-effective, and easily implemented solution for safety monitoring and early detection of abnormal flight conditions in agricultural UAV delivery.