Loading...
Advancing ecological survey through automation, artificial intelligence and unmanned aerial system
Date
2025
Abstract
Ecological surveys are critical for conservation and the development of renewable energy, particularly in assessing biodiversity and environmental impacts. However, traditional methods face challenges in scalability, accuracy, and resolution. This thesis addresses these issues through automation and high-resolution un-manned aerial vehicle (UAV) systems.
The first part focusses on automating workflows for digital aerial surveys in an offshore wind energy context. A convolutional neural network-based system was developed to automate data screening, reducing processing times from months to hours. Validation over 15 months of survey data showed a 36% increase in marine mammal detection rates compared to manual methods. Bird detection matched human accuracy but highlighted limitations in current survey resolutions, emphasising the need for higher-quality data.
The second part addresses these limitations by improving image resolution and transitioning to UAV-based systems. Extensive trials explored thermal imaging and RGB cameras in diverse environments, from offshore, intertidal, cliff areas, to onshore trials. Although thermal imaging faced challenges classifying birds, high-resolution RGB systems allowed for accurate classification, but at the cost of field of view. This led to the development of a Modular Detection and Targeting System (MDTS), integrating thermal and RGB imaging for real-time ecological monitoring. Field trials validated its scalability and adaptability for avian and mammalian species
By combining automation, high-resolution imaging, and UAV adaptability, this research contributes novel solutions to ecological surveying. The findings have significant implications for the acceleration of offshore wind energy development, the enhancement of wildlife monitoring, and the advancement of the use of drones in ecological research. Future work should focus on improving sensor technologies and integrating automated systems across traditional and UAV platforms.
Supervisor
Dooly, Gerard
Trslić, Petar
Santos, Matheus
Trslić, Petar
Santos, Matheus
Description
Publisher
University of Limerick
Citation
Collections
Files
ULRR Identifiers
Funding code
Funding Information
Sustainable Development Goals
External Link
License
Attribution-NonCommercial-ShareAlike 4.0 International