Drone for inspecting solar panel fields

The problem

Solar fields in Jordan have grown, yet inspecting them is still manual: a technician walks between rows checking panels by eye. A dusty or cracked panel doesn't stop working; it just produces less, so there's no obvious fault to notice, and the result is a hidden, cumulative loss. A drone can cover in one hour an area that takes a full day on foot, and can spot patterns the naked eye misses.

How it works

1. PlanningThe field's boundaries are set on the map, and the system generates a flight path that covers the whole field.
2. FlightThe drone follows the path at a constant altitude using the pressure sensor and GPS.
3. ImagingThe camera photographs every row of panels, recording the coordinates of each image.
4. AnalysisThe system analyzes the images and classifies each panel as clean, dusty, shaded, or broken.
5. LinkingEach observation is linked to its coordinates, pinpointing the panel in question.
6. ReportThe dashboard shows a field map colored by panel condition, with a list of maintenance priorities.

Project features

• It saves real inspection time, measurable in hours and dinars.
• It combines programmed flight, image analysis, and geolocation.
• It belongs to a growing sector in Jordan, and the idea has a real market use beyond the academic setting.
• The output is a clear visual report that's easy to present to the defense committee.
• It can be extended with a thermal camera to detect hot spots.

Expected challenges

These are the main obstacles to expect during the build, laid out clearly so you know them before you start.

• Keeping altitude and path steady in the wind over open fields.
• Sunlight reflecting off the panel saturates the image and hides details.
• Telling dust from shade and from cracks in the image is the hardest technical part.
• Balancing the flight time the battery allows against the field's size.
• Linking each image to coordinates precise enough to identify a specific panel.

The consultation gives you a solution for each one.

Expected committee questions

Prepare your answers in advance: these questions come up often in defenses of similar projects.

1. How was the model trained to detect dusty and broken panels, and where did the training images come from?
2. How is each image linked to the correct panel number in the field?
3. How do the sun's angle and reflections affect the analysis accuracy?
4. What accuracy did the model achieve, and how was it measured?

Practice the defense, question by question