Tahr Defender: An Artificial Intelligence-Powered Drone System for Arabian Tahr Conservation

Conserving endangered wildlife like the Arabian Tahr involves significant challenges, such as ensuring ranger safety and monitoring remote habitats effectively. The Tahr Defender project addresses these challenges by integrating drone technology, Artificial Intelligence (AI), and a mobile application into a comprehensive conservation tool. A custom dataset, including Arabian Tahr and intruders, was developed by combining public datasets with data collected during field visits using drones specifically for the Arabian Tahr, as no public dataset exists for this species. Utilizing the You Only Look Once (YOLOv11) object detection model, the system identifies Arabian Tahr and potential intruders with accuracies up to 91 %. A Contrastive Language-Image Pre-Training (CLIP) model further enhances the system by classifying habitats and providing ecological insights. A mobile app, backed by a server and database, was developed to enable rangers to view detections, track wildlife activity, and manage flight data efficiently. This app ensures seamless integration of AI-driven insights and drone technology into a user-friendly platform. The system provides an innovative approach to not only improving ranger safety and operational efficiency but also offers a scalable and practical solution for monitoring endangered species and supporting wildlife conservation efforts.