From Fire Detection to Forest Recovery: An Integrated UAV-IoT System for Wildfire Monitoring and Reforestation

Wildfires and the degradation of forest land pose a major threat to ecosystems, particularly in Northern Thailand, where the dry season brings recurrent fires, severe haze and long-term biodiversity loss. As part of Project FireFly, this work presents an integrated UAV-based approach spanning the full cycle from the early detection of forest fires to the restoration of burnt landscapes. For fire monitoring, low-cost IoT ground modules measure temperature, humidity and smoke beneath the canopy. A LoRa mesh network has been implemented for communication between the distributed sensor nodes. As 4G coverage is limited in remote forest areas, a multirotor UAV carrying a 4G-enabled air module acts as a communication relay, collecting sensor data along pre-programmed survey routes while simultaneously surveying the area for fire using a thermal infrared camera. Data is streamed to an online dashboard, where machine learning models provide fire probability estimates for emergency response services. The same UAV-IoT infrastructure is extended beyond fire detection towards biodiversity monitoring and assessment of forest restoration in areas destroyed by fire. Additional sensors enable wildlife surveillance in remote areas, while aerial surveying of tree health and canopy structure quantifies the carbon sequestration potential of reforested areas. For forest restoration, a UAV-based seed-dropping system has been developed as part of Project FireFly to enable direct seeding of native tree species. A lightweight carriage mechanism has been integrated with the multirotor platform to release individual seed balls at pre-defined locations. Field validation near Chiang Mai, including trials at a restored forest site operated by the Forest Restoration Research Unit at Chiang Mai University, demonstrates the feasibility of the proposed approach combining fire detection, biodiversity monitoring and UAV-based reforestation. The work illustrates how UAV infrastructure developed for wildfire response can be repurposed to support the longer restoration cycle that follows a fire.