From Inspection to Measurement: Why Solar Data Needs a Rethink

Modern solar infrastructure is built on precision. But the way it is often inspected still leaves too much room for variability.

Across the industry, it is not uncommon for two images of the same solar module, captured under the same measurement conditions, to produce different analytical results. The panel has not changed. Yet the outcome does. That inconsistency points to a deeper issue: many inspection workflows are still not true measurement systems. They are interpretations of visual data.

“Without a stable input, the numbers you extract, the power loss you calculate, even your whole analysis becomes subjective.”

— Karl Bedrich, CTO, Quantified Energy

The challenge with unstable inputs

At its core, inspection depends on the integrity of input data. But in practice, that input is highly variable. Differences in camera calibration, perspective, image sharpness, noise patterns and even image scaling can significantly affect what is ultimately measured.

These variations may appear subtle in isolation, but their impact compounds across large portfolios. Metrics such as defect detection and power loss estimation, which are expected to be objective, become sensitive to how the data was captured rather than what the data represents. In such conditions, analysis begins to drift from measurement towards interpretation.

A structural limitation in conventional imaging

The challenge is further complicated by the nature of the signal itself. Electroluminescence, or EL, which is used to identify defects in solar modules, exists in the near-infrared spectrum, just beyond visible light.

Most conventional imaging systems rely on silicon-based sensors such as CCD or CMOS, which perform well in visible wavelengths but lose sensitivity in the near-infrared range. To compensate, traditional inspection methods often depend on long exposure times and controlled, static environments. While effective in laboratory settings, these approaches struggle to translate to large-scale, real-world operations.

The result is a fundamental mismatch between the physics of the signal and the capabilities of the tools used to capture it.

“What comes out isn’t a photograph anymore. It’s a quantitative measurement.”

— Karl Bedrich, CTO, Quantified Energy

Reframing inspection as a measurement problem

Addressing this mismatch requires more than incremental optimisation. It calls for a reframing of the problem itself.

Rather than asking how to capture better images, the more relevant question is this: how can inspection be treated as a measurement system?

This shift in perspective is what underpins Quantified Energy’s approach. Instead of relying on conventional imaging pipelines, the system is designed around a more integrated view of sensing, calibration, data processing and flight execution.

1. Align the sensor with the signal

Using InGaAs sensors, which are inherently sensitive in the near-infrared range, makes it possible to capture electroluminescence with significantly stronger signal strength. This reduces the need for long exposures and enables imaging under motion, which is essential for field deployment.

2. Move from single images to reconstructed data

Single-frame imaging is inherently limited by noise and resolution constraints. A multi-frame approach, capturing dozens of frames per second, allows for geometric correction, radiometric calibration and precise alignment at a subpixel level. Through signal averaging and outlier rejection, it becomes possible to reconstruct a higher-fidelity representation of the underlying data.

In this context, the output is no longer simply an image. It becomes a derived measurement.

3. Reduce variability at the point of capture

Human-operated inspection introduces variability that is difficult to standardise, particularly under field conditions. By integrating autonomous flight control and real-time optimisation, where parameters such as flight speed are dynamically adjusted based on signal quality, the system reduces dependence on operator judgement.

The role of the operator shifts from manual execution to supervision. That is an important transition, because scalability depends not only on technical accuracy, but also on operational repeatability.

“When your data is stable, repeatable and captured at scale, inspection stops being an art. It becomes a simple measurement.”

— Karl Bedrich, CTO, Quantified Energy

From art to infrastructure

Taken together, these changes represent a broader transition. Inspection moves away from being a specialised, skill-dependent activity and towards becoming an infrastructure layer, one that produces consistent, repeatable data across assets and over time.

This distinction is not merely technical. It has implications for how solar assets are managed, evaluated and financed. In large-scale energy systems, reliability is not defined by occasional accuracy, but by consistency.

Why this shift matters now

As solar deployments continue to scale globally, the limitations of traditional inspection methods become more pronounced. Operators require not only visibility into current performance, but confidence in how that performance is measured. Investors and stakeholders, in turn, depend on data that is stable and comparable across time and geography.

In this context, inspection is no longer a peripheral activity. It becomes central to asset integrity, operational decision-making and long-term value creation.

Towards a more measurable future

The evolution of solar inspection reflects a broader pattern seen across industries: as systems scale, measurement replaces observation.

The question is no longer whether defects can be detected, but whether they can be quantified in a way that is repeatable, defensible and actionable. Solving this requires alignment across sensing, data processing and operational execution. It requires treating inspection not as a visual exercise, but as a measurement discipline.

Ultimately, better decisions begin with better data.

Quantified Energy is a Singapore-based deep-tech company focused on improving solar asset performance through AI-driven inspection and analytics. Its core solution uses autonomous drone electroluminescence mapping to deliver lab-quality field inspections, enabling earlier defect detection and more reliable asset assessment.

Learn more at quantified-energy.com