Leveraging Drone-Assisted Surveillance for Effective Forest Conservation: A Case Study in Australia’s Daintree Rainforest

Nowadays, there is global consensus on the threats to forests and their crucial role in mitigating global warming and its impact on Earth’s biodiversity. Both private and public entities, alongside governments, have engaged the most advanced technologies to safeguard and monitor forests against encroachment. This paper examines the application of various drone technologies in the surveillance of forest areas. The system described herein employs drones to continuously survey forests, recording any changes, particularly in instances of encroachment or fire. The data captured are transmitted to a control unit for subsequent analysis. To circumvent the risk of task failure due to technical challenges, monitoring tasks within a predefined flight duration are allocated to the available drones. Given the critical nature of timing in the success of these tasks, this study addresses the forest monitoring challenge by seeking to minimize the maximum time required to complete all monitoring tasks. This challenge was addressed through the development of a suite of enhanced algorithms aimed at optimizing task efficiency. The primary goal of the proposed methodology is to afford the monitoring system additional time, thereby enabling the handling of an increased volume of tasks and providing support to firefighting teams in responding to forest fires. The system’s adaptability to new, unforeseen forest fire scenarios through the generation of novel solutions is also discussed. Extensive testing involving 1,350 different scenarios has demonstrated the effectiveness of the proposed algorithms in reducing the maximum time needed for the completion of surveillance tasks by drones. The most effective algorithm was the two-group clustering algorithm (TGC), which achieved a success rate of 97.2%, with an average gap of less than 0.001 and an average computation time of 0.016 seconds. Furthermore, the application of this methodology to a case study of the Daintree Rainforest in Australia showcases the potential and real-world applicability of the proposed system, highlighting its performance and adaptability.