Simulation-Based Optimization of Drone-Assisted Search and Rescue Operations in a Flood-Prone Area

Abstract: Flooding is one of the most common and dangerous hazards in the Philippines, especially in the low-lying communities such as Purok 4, Barangay San Roque, Mercedes, Camarines Norte. Using the traditional search and rescue (SAR) operations often become slow and unsafe because the rising water levels, poor visibility, and floating debris make the movement difficult for the rescuers. This study explores the use of the drone-assisted navigation to support the rescue operations during the floods. Using the Agent-Based Modeling (ABM) in a NetLogo, the simulation can make the rescuers, victims, and drones interacting within a simplified map of the barangay. There are four navigation approaches that were tested: No Algorithm (random movement), Dijkstra’s Algorithm, Greedy Best-First Search, and a Hybrid method combining Greedy detection with Dijkstra routing.Multiple trials were conducted to measure the detection time, rescue time, and the total operation time. All scenarios were able to rescue the victims, but their performance is different depending on flood levels and environmental difficulty. The No Algorithm approach was only effective in the simple situations, while Dijkstra produced a reliable and safe routes in the moderate flooding. Greedy also allowed a fast initial detection but struggled in complex environments. The Hybrid method showed the best adaptability to challenging conditions. Overall, the study highlights how combining the speed and safe pathfinding can improve the drone-assisted rescue operations in real flood responses.