A Machine Learning Based Synergistic Swarm Drone System to Eradicate Locust Swarms: Project Mynah

The United Nations Sustainable Development Goal (UNSDG) 2 - calls for creating a world with zero hunger by 2030. However, locust swarms threaten over 10% of the global food supply. Locusts normally live in a harmless solitarious state. However, on overcrowding multiple morphological transformations of the solitarious locusts turn them into a gregarious swarm forming state. Due to climate change locust swarms are predicted to become more frequent, intense and long-lasting. The current detection, monitoring and control systems for locusts is lacking. The developed Project Mynah ensures reliable regular locust hotspot monitoring, early gregarization detection and an effective eco-friendly control which induces the solitarization of the locusts, preventing swarm formation. The solution consists of an AI system including a Yolov8 nano object detection model and a Mobile Net V2 CNN which accurately detects and differentiates between solitarious locusts and gregarious locusts based on morphology. To facilitate reliable and regular monitoring of remote areas, the system includes a synergistic autonomous swarm of drones governed by bioinspired decentralized algorithms like vulture frontier radial exploration and A*. The overall system has been engineered to optimize resource usage and ensure global scalability.