Using UAV to Detect Solar Module Fault Conditions of a Solar Power Farm with IR and Visual Image Analysis

Round 1

Reviewer 1 Report

The authors should clearly state the novelty of the presented approach and emphasise the differences between their work and other works presented in [2] and [6] for example.

The final result is a cumulative distribution function which still has to be examined by maintenance staff. Shouldn't the original thermal and IR images be sufficient as well?

This approach would have more sense if it would serve as an input to some kind of an artificial neural network. The machine learning techniques could then be employed to discern the normal vs. faulty panels.

The work presented in this paper seems to be a follow up on [8]. That article was of a significantly higher quality in terms of novelty and scientific merit.

There are several grammatical errors in the text, please have it checked for language and style.

The textboxes in CDF plots should be removed, they are placed directly over the curve rendering it useless.

Overall, the paper is of average quality and needs to be improved. The scientific novelty is unclear and questionable. I recommend a major revision.

 

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Reviewer 2 Report

The paper is the result of quite useful experiment showing scientific quality.

The conclusions: "The advantage of the innovative method proposed in this paper is to use an infrared imaging system with drone functions to detect faults in the solar power system. And it uses MATLAB image analysis software to analyse the IR image."

This is not a novelty but common experiment. You need emphasise more your input.

Gaussian filtering is one of more treatment methods of IR images. You should discuss its advantages/disadvantages  and eg compare with other alternatives.

Some recommendations/comments (some of many others)

UAV in the title and abstract - this abbreviation is not generally known and should be explained

Title 2. Materials and Methods - here "Materials" is not very convenient to describe the matter

p.3/l.7 - "replace cell" - is that possible or do you mean replace the module?

You mention deviations. Is that absolute deviation? Usually deviation is plus/minus

Language need to be corrected. There are more imperfectness like eg p.14/l.5:

"However bright areas are more visible than IR images" (eg.: Bright areas after filtering can be better recognized than at IR images)

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Reviewer 3 Report

The introductory part and abstract does not justify properly the proposed paper, even in the discussion of results.

The UAV system configuration was not presented.

The sample dataset used in this paper was not properly detailed.

There's no comparative analysis to justify the claim of this paper.

The references used are not enough to establish this paper's objectives.

Provide latest and more related references.

The histogram and cumulative chart presented do not justify fault conditions because there's no comparative analysis used between a non-fault thermal image.

This paper uses UAV system, there's no analysis or presentation of why UAV is appropriate for fault detection.

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.doc

Round 2

Reviewer 1 Report

The authors have improved their manuscript according to the reviewers' comments. It can be accepted for publication in this version.

Author Response

Thank you! Your approval is highly appreciated.

Reviewer 2 Report

The paper was properly improved and its message one can find useful, either technical or scientific.

Author Response

Thank you! Your approval is highly appreciated.

Reviewer 3 Report

Additional technical keywords are necessary.

Minimize the conclusion contents. Some of the statements should be presented in the discussion of results.

In analysing figure 11 image, provide justification on the histogram and cumulative chart that it is a defective solar module.

In the presented sample images, especially in Figure 14, how can the reader identify that the flare in Figure 14.a, then to c, to d. that this is not just a reflection of light or any configuration of the environment, provide the technical specification of this testbed to support the presented image.

Author Response

Please see the attachment.

Author Response File: Author Response.doc