Autonomous Drone Inspection For Solar PV Parks With AI
Problem Statement:
The maintenance of Solar PV parks faces significant challenges due to environmental factors, such as weather erosion, obstructions like bird droppings and dust, and the technical issue of localized heating in photovoltaic modules. These problems lead to reduced efficiency, high safety risks, and substantial maintenance costs and workload, underscoring the urgent need for an innovative solution to streamline and improve the inspection and maintenance process.
Aim:
Develop a drone prototype to obtain the temperature in the Solar PV parks.
Objectives:
1. To design an autonomous drone system for temperature measurement.
2. To develop an IoT dashboard for monitoring
3. To optimize the PV system performance using data analytics
Presentation Slides
While challenges with the FLIR Lepton camera hindered the first objective, thermal image datasets were trained using YOLOv8. Hotspot detection was achieved in the dashboard, evaluated for field performance and accuracy.
Despite limitations, the project provided invaluable real-world problem-solving experience. The faulty thermal camera posed integration challenges, impacting the temperature detection goal. Unattained drone flights hindered outdoor thermal image capture. Refinement is suggested, including exploring alternate camera models and conducting real-world testing at PV farms for comprehensive assessment.
Additionally, the dashboard could feature temperature data display and waypoint selection for drone navigation. The technology’s potential extends beyond PV parks, with applications in wind turbine farms, enhancing inspection safety. Collaboration with utility technicians remains crucial for accurate fault identification. Although challenges persist, the project demonstrates progress in autonomous drone-based inspection, emphasizing the importance of innovation, collaboration, and continuous improvement in renewable energy systems.