Abstract

Agriculture faces growing challenges, including rising costs, labor shortages, and the increasing need for efficiency and product quality. The Internet of Agricultural Things (IoAT) offers promising solutions through technologies such as weather and soil sensors, automated feeders, RFID tags, and computer vision systems, enabling data-driven decision-making. However, widespread adoption is hindered by constraints such as limited broadband connectivity, low-power device capabilities, harsh environmental conditions, and farmer acceptance of new technologies.

In this talk, we first provide an overview of IoAT technologies and their impact on modern farming. Motivated by connectivity challenges, we then introduce a UAV-based framework for high-precision crop monitoring that leverages statistical inference and machine learning to minimize operational costs. Additionally, we present iCrop+, a LoRa-based crop disease detection system that integrates on-device AI with deep learning for remote processing. iCrop+ intelligently offloads data via LoRa communication using a combination of category-based and sample-based offloading strategies to optimize accuracy and efficiency. We evaluate both approaches using real-world prototypes, demonstrating their viability for deployment in resource-constrained agricultural environments.