Real-time hydroponic plant disease detection using YOLOv11n on edge devices

Hydroponic farming offers a sustainable alternative to traditional agriculture but is highly prone to rapid disease transmission due to its shared water systems. Timely detection of plant diseases is critical to prevent widespread crop loss. In this research, the YOLOv11n object detection model was evaluated in detail for the purpose of real-time hydroponic plant disease detection, and its accuracy, inference speed, power consumption, and resource utilization were compared through various edge devices such as Raspberry Pi 5 (Raspberry Pi Holdings, Cambridge, UK), NVIDIA Jetson Nano (NVIDIA Co., Santa Clara, CA, USA), and AMD Radeon Vega 8 (AMD Micro Devices, Inc., Santa Clara, CA, USA). The results not only confirm the applicability of light-weighted and quantized deep learning models for planting disease detection at early stages in controlled hydroponic environments but also give practical knowledge for hardware-model trade-offs in the context of sustainable edge AI-based smart farming systems.