A LoRaWAN-BLE Based AIoT Smart Farm Management and Control System

The indoor vertical farming is a mature solution for Smart agriculture for Urban farming applications. However, it requires enhancement for agricultural productivity. This paper presents a comprehensive LoRaWAN-BLE based Artificial Intelligence of Things (AIoT) smart farm management and control system designed to automate farm operations with minimal human intervention. The system integrates BLE nodes interfaced with various sensors and actuators to monitor and control environmental conditions. LoRaWAN-based cameras capture images of crops, enabling two advanced machine learning models to assess plant growth. The Effi-Att-UNet3+ model assesses crop coverage with a pixel accuracy (PA) of $\mathbf{9 2. 8 \%}$, and the BiFPN-YOLOv8 model evaluates plant maturity through leaf segmentation and shape detection with mAP@50 of 0.743 and 0.699, respectively. The proposed system effectively transmits sensor data via a LoRaWAN network, ensuring long-range communication with low power consumption. This smart farm system not only digitizes the growing process, making it more accessible and efficient for farmers, but also significantly reduces manual labor, contributing to enhanced agricultural sustainability and food security.