Drone‐Based Solar Cell Inspection With Autonomous Deep Learning
Bibliographic record
Abstract
Regular inspection and maintenance are crucial for ensuring the optimal performance of solar panels. However, conventional manual methods can be laborious, time consuming, and expensive, especially for large and inaccessible installations. Aerial inspection has the potential to overcome these limitations and improve operational flexibility. To fully leverage the potential of aerial inspection, we present a summary overview of drone-based photovoltaic module inspection and a case study demonstrating the integration of autonomous navigation and machine learning techniques for defect detection. In the case study, a convolutional neural network ( CNN ) based framework that can autonomously detect defective solar cells using aerial robots is integrated with the autonomous navigation of the aerial robot. There are two main phases for this framework: detection of the solar panel location and identification of the solar cell defect with a feasible set of trajectories. The solar panel is identified with a shape detection algorithm and the defects are classified using electroluminescence ( EL ) images with a CNN, based on the VGG16 architecture; various approaches to avoid overfitting are presented to achieve better performance. Seven solar cell states can be detected including breaks, finger interruptions, material defects, and microcracks. This pipeline is demonstrated virtually on the NEST building at the Swiss Federal Laboratories for Materials Science and Technology. The research hub is reconstructed in AirSim with real data and it is shown that the aerial robot can detect the location, conduct an approaching trajectory, extract the image, and transfer it to the CNN for defect identification.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; both teacher heads agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".