Design and Implementation of an NB-IoT-Based Grain Warehouse Environmental Monitoring System

An NB-IoT-based grain-warehouse environmental monitoring system is designed to improve the timeliness, coverage, and abnormal-event handling of conventional manual inspection. An STM32F103C8T6 integrates DHT11 temperature/humidity, MQ-135 gas, VOC/CO2/formaldehyde, and BH1750 illuminance sensors. Data are uploaded through NB-IoT using MQTT, while an OLED, buzzer, Web client, and mobile application provide local and remote display/alarm functions. The terminal implements data validation, filtering, threshold decision, reconnection, and retransmission. In simulated tests, connection success rates under normal and weak signals were 99.6% and 96.2%; 1 KB latency was 0.42 s, 1 h packet loss was 0.05%, and recovery after simulated interruption was 8.1 s. These results validate the single-node communication and functional closed loop under the tested conditions, but not reference-grade sensor accuracy, long-term warehouse robustness, multi-node capacity, or battery lifetime. Additional calibration, field, scalability, and energy tests are therefore required.