Enhancing Crop Yield and Sustainability through an AIDriven Pest Monitoring and Decision Support System

Agricultural industry which is a large sector of the global economy continues to be struck by pest menaces which destroy crops and fill the economy with uncertainty. This particularly is crass to small farmers in such places as India. Old fashioned pest checks are a pain in the neck–they work too much, are too slow, and most of the time they miss the target until it is too late. It signifies that we are spraying chemicals all over and it is not economical as well as not environmentally friendly. We, therefore, addressed these issues through a project named as Smart-Crop-Monitor. It is an expensive automated system that detects and identifies pests in real-time. We trained the most recent YOLOv8 framework, which had been refined on 12 types of pests of a large and diverse dataset. It needs a full site, written in custom code. FastAPI is the in the background, managing data, on the front, a React interface makes it easy and enjoyable to see and interact with photos and videos. We designed the system to run on cheap edge hardware such as an old laptop or a single board computer in addition to a modular alerts system that can buzz you immediately through whichever channel you prefer. When we were on the validation, our YOLOv8n achieved 0.444 mAP50-95, which is something like, it is quite consistent and can be a practical tool. Overall, this paper provides a complete design of a hand-to-hand smart farming kit, demonstrating that the integration of the latest AI, web technologies, and IoT can be scaled up to make the pest control data-driven and contribute to the advancement of sustainable farming.