FIDAL

The DADOLTI project proposes an advanced, drone-based solution for inspecting overhead power lines and towers. It leverages a unique Hierarchical Inference (HI) framework, combining lightweight deep learning (DL) models deployed on drones with more comprehensive models hosted on nearby edge servers. This setup ensures real-time decision-making without requiring continuous cloud data transfers. Using the Patras5G/PNET infrastructure, the project’s trials will optimise response times and reduce operational costs through targeted deployment of 5G-enabled drones that seamlessly switch between local and edge processing based on network conditions and confidence levels in the drone's analysis. DADOLTI addresses multiple technical and social challenges by enhancing predictive maintenance capabilities, reducing manual inspection needs, and enabling safer, eco-friendly inspections. It promises significant benefits for distribution system operators (DSOs), drone providers, and the larger edge ecosystem, supporting rapid fault detection and contributing to the reliability and sustainability of energy distribution systems. The project also prioritises privacy and security, ensuring GDPR compliance and safeguarding data throughout the inspection process.