PV Module Manufacturing Quality Control
AI-Powered Electroluminescence (EL) Testing
Verifying photovoltaic module quality requires more than identifying visible defects. Hidden cell defects can compromise long-term reliability, while manually reviewing large volumes of Electroluminescence (EL) images is often impractical.
STS AI-Powered EL Testing helps buyers independently identify hidden manufacturing defects, verify compliance with project-specific acceptance criteria and make informed shipment decisions before modules leave the factory.
AI-Powered EL Testing
AI-Powered Electroluminescence (EL) Testing
EL Testing
Electroluminescence (EL) Testing is an industry-standard quality control process used during PV module manufacturing to identify hidden defects that cannot be detected by the human eye.
The Need for AI
Before AI-enabled analysis, independent verification was typically limited to just 1–5% of production, because manually reviewing every EL image was impractical.
How AI is Used Today
AI now makes it practical to evaluate EL results for up to 100% of production. Manufacturers increasingly use AI as part of their quality control process, while buyers can either rely on manufacturer analysis or engage independent experts like STS to independently verify results against their project-specific quality requirements.
Preventing Avoidable Costs
Preventing Avoidable Costs
Independent Verification can Prevent Avoidable Module Replacement Costs
Manufacturers are contractually required to evaluate modules against project-specific quality requirements.Â
However, buyers may have limited visibility into how manufacturer AI is configured or whether those requirements have been applied consistently across the full production run. Independent verification provides a second layer of assurance by evaluating EL results against the buyer’s
project-specific defect definitions, acceptance criteria and quality thresholds before shipment.
On a typical project, independent verification of EL results can help reduce replacement costs and avoid delays associated with removing and replacing non-conforming modules.
Why Choose STS
The Difference Isn't AI. It's How STS Applies It.
Many providers now use AI to accelerate EL image analysis.
STS goes further by configuring every inspection around the buyer’s project requirements and training the AI model specifically for each project’. These STS are, to the best of our knowledge, are not currently offered by other providers and help identify additional non-conforming modules before they create replacement costs, delays and rework.
| Typical AI EL Providers | STS |
|---|---|
| Generic AI model | Project-specific AI training |
| Standard defect library | Buyer-specific acceptance criteria |
| Automated defect classification | Expert-verified defect classification |
Why STS Is Different
DISTINCTIVE STS CAPABILITY​
Configured for Your Project​
Every inspection is configured around your project's defect definitions, acceptance criteria & quality thresholds, not generic manufacturer defaults.​
DISTINCTIVE STS CAPABILITY​
Project-Specific AI Training​
STS trains the AI model using project- specific EL images and customer requirements to improve consistency from the first module inspected.​
AI + Manual Validation​
Potential defects are manually verified by STS experts, providing an additional layer of quality assurance.​
High-Throughput Inspection​
Analyze more than 150,000 EL images per week with reporting available within 48 hours after receiving the complete EL image dataset.​
PV Module Defects & Impact
Hidden Manufacturing Defects AI-Powered EL Analysis Identifies
Cracks & Cell Damage
Branching or multidirectional cell cracks visible in EL images that may disrupt electrical pathways and increase the risk of power loss and further cell damage.
A linear crack extending across a solar cell that may interrupt current flow and develop into electrically inactive areas under mechanical or thermal stress.
Surface or cell damage identified during EL inspection that may indicate manufacturing or handling issues and potential weakness within the PV cell.
Physical damage at the corner of a solar cell, typically associated with manufacturing or handling, which may affect cell integrity and electrical performance.
Abnormal cell patterns or imprints associated with manufacturing fixtures or the sintering process that may indicate inconsistencies during PV cell production.
Module Manufacturing & Assembly
Electrical & Performance
An abnormal electrical condition within a solar cell that can reduce its effective contribution to module output and indicate a potentially significant manufacturing defect.
Defects or interruptions affecting the cell’s conductive grid that may restrict current collection and negatively influence PV module electrical performance.
A portion of the solar cell showing little or no electroluminescence, indicating an area that is not contributing normally to electrical current generation.
Abnormalities associated with bypass diode operation that may affect current flow, module protection, and overall PV module performance under operating conditions.
A localized area of unusually high EL intensity that may indicate abnormal electrical behavior or manufacturing variation requiring further assessment.
A distinct dark linear feature in an EL image that may indicate interrupted electrical activity, cell damage, or another manufacturing-related non-conformity.
Abnormal EL patterns along the short edge of a cell or module that may represent imaging artifacts, manufacturing inconsistencies, or localized defects requiring classification.
Why Defects in PV Modules Matter
Even minor manufacturing defects can become increasingly costly as PV modules move from production to operation. Identifying non-conforming modules before shipment is the most cost-effective opportunity to protect project performance, schedules and long-term asset value.
Hidden defects enter production before modules leave the factory.
Microcracks, soldering defects, inactive cell areas and other manufacturing irregularities are often invisible during standard visual inspection. Electroluminescence (EL) testing can reveal these hidden electrical defects, but the effectiveness of the review depends on how consistently the images are analyzed and what acceptance criteria are applied.
The cost of correction increases significantly once modules are shipped.
Modules approved during manufacturing may be transported across continents before arriving at the project site. Replacing these non-conforming modules will require additional work, creating avoidable schedule disruptions and added costs.
Manufacturing defects can become operational problems over time.
Some defects remain stable, while others may worsen due to transportation stresses, thermal cycling, mechanical loading and long-term field exposure. As degradation progresses, modules may experience reduced power output, accelerated aging or increased likelihood of premature failure.
Small manufacturing issues can have long-term commercial consequences.
Undetected defects can contribute to lower energy production, warranty claims, module replacement programs and increased operational risk over the life of the project. Identifying non-conforming modules before shipment helps reduce these risks before they become significantly more expensive to address.
Why AI Improves EL Testing
Why AI Improves Electroluminescence (EL) Testing
As PV module production scales, manufacturers generate thousands of Electroluminescence (EL) images that must be reviewed for hidden manufacturing defects. AI-powered EL analysis enables rapid, consistent review of large image datasets, helping buyers identify potential manufacturing risks more efficiently while supporting independent technical verification.
Â
| Traditional EL Review | STS AI-Powered EL Testing |
|---|---|
| Manual review of selected EL images | Rapid analysis of large EL image datasets |
| Time-intensive | High-throughput image analysis |
| Reviewer-to-reviewer variability | Consistent defect classification |
| Difficult to scale | Flexible inspection scope, including up to 100% of production |
| Slower reporting | Faster reporting with AI and expert validation |
AI doesn’t replace technical expertise; it enables STS experts to review more EL images with greater speed, consistency and confidence.
- Analyze significantly larger image datasets​
- Reduce reliance on labor-intensive manual review
- Faster inspection and reporting
- Improve defect classification consistency
- Flexible inspection scope, including up to 100% of production where required
- Scale efficiently for high-volume PV module manufacturing
How STS AI-Powered EL Testing Works
How the STS AI-powered Electroluminescence Testing Service Works​
International PV Standards
Aligned with International PV Industry Standards
STS performs AI-powered Electroluminescence (EL) Testing in accordance with internationally recognized photovoltaic (PV) standards and buyer-defined technical requirements. By combining industry guidance with independent technical expertise, STS delivers consistent, objective evaluation of PV module manufacturing quality.
IEC 61215
PV Module Design Qualification
Defines the design qualification and type approval tests used to evaluate the long-term performance and reliability of crystalline silicon PV modules.
IEC TS 60904-13
Electroluminescence (EL) Imaging
Provides guidance for acquiring, interpreting, and evaluating EL images to identify and classify hidden PV cell and module defects.
IEC TS 62446-4
PV Module Quality Assessment
Establishes best practices for evaluating PV module quality using imaging techniques during manufacturing and field inspections.
Deliverables
Deliverables
Frequently Asked Questions
Frequently Asked Questions
Electroluminescence (EL) Testing is a non-destructive inspection method that uses infrared imaging to identify hidden defects within photovoltaic (PV) cells and modules that cannot be detected through visual inspection alone. By identifying manufacturing defects before shipment, EL Testing helps manufacturers, developers and buyers reduce quality risks, improve procurement decisions and support long-term PV module performance and reliability.
AI-powered Electroluminescence (EL) Testing can identify a wide range of hidden PV cell and module defects, including tree-shaped cracks, single-direction cracks, soldering defects, inactive areas, grid defects, bright spots, black lines, dark edges, failure diodes and other manufacturing non-conformities. Identifying these defects before shipment helps reduce manufacturing, warranty and long-term performance risks.
Traditional EL inspection relies on manual review of selected EL images, which can be time-consuming and difficult to scale consistently. STS combines AI-powered image analysis with independent technical expertise, enabling rapid analysis of large EL image datasets while maintaining consistent defect classification. AI identifies potential non-conformities, and STS experts independently verify representative results before findings are reported, providing greater confidence in inspection outcomes.
Yes. STS offers flexible inspection scopes ranging from targeted sampling to review of up to 100% of production, depending on project requirements and available EL image datasets. AI rapidly analyzes the selected images, while STS experts independently verify representative AI findings before issuing the final inspection report, combining the efficiency of AI with independent technical oversight.
Yes. STS applies buyer-defined acceptance criteria based on each project’s technical and quality requirements rather than relying solely on manufacturer-defined thresholds. To perform the analysis, manufacturers typically provide the EL image dataset together with relevant production batch and project information required to complete the inspection and reporting process.
STS performs AI-powered Electroluminescence (EL) Testing in alignment with internationally recognized photovoltaic (PV) industry standards, including IEC 61215, IEC TS 60904-13, and IEC TS 62446-4, where applicable. These standards support consistent EL image interpretation and independent quality assessment as part of PV module manufacturing and pre-shipment inspection programs.