Edge Computing for AI-Driven Acoustic Detection of Birds Pest in Senegalese Rice Paddies

In Senegal, rice cultivation plays a key role in food self-sufficiency and rural development. However, several obstacles stand in their way: pests, four-legged animals and, above all, harmful birds (quelea quelea). These birds are responsible for colossal yield losses. Farmers use archaic methods to deter them, which are often ineffective or costly, and do not provide constant protection. Faced with this endemic problem, an innovative technological solution is proposed, combining the Internet of Things (IoT), Artificial Intelligence and Edge computing. The system is based on a network of microphones capable of capturing sound anywhere in the rice paddy, and above all, continuously. Bird songs can be specifically identified thanks to real-time analysis of the recorded sounds made possible by on-board AI models. This solution allows for very low latency and fast, autonomous detection without requiring a constant internet connection. The system assists in bird deterrence by quickly and accurately identifying their presence. This intelligent system helps to reduce crop losses, sustainably improve agricultural productivity and enhance precision farming.