Feasibility and Operational Limits of a Minimum-Cost Indirect UAV Thermal Sensing Workflow Based on Smartphone-Displayed Infrared Video

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The topic is interesting, whereas the text is written in proper English, it is informative.

Suggestions and comments:

1) In the Authors and Affiliation section check how to properly use Capital letters in case of your Institutions and make necessary corrections.

2) The technical specs of the drone, smartphone and camera should be summarized on the form of a table.

3) There seems to be a change in font size/type from time to time. As an example, look at lines 233-237 on page 7. A simmilar change seams to appear in section 2.9 on page 11.

4) Line 274 seems to be missing a beginning.

5) The paper lacks a proper Related Works section, summing up previously published works on the aforementioned topic. Do summarize advancements in either drone technology, thermal inaging, sensors/cameras, etc.

6) The title includes the word smartphone, however very little attention is given to this mobile device.

7) Check the font size and type in case of tables and make necessary corrections according to the Journal template.

8) After reading the whole paper I cannot notice the main aim of it. Was it the integration of a consumer drone + smartphone + camera? Was it the analysis of payload vs performance of the drone? Was it the stability or accuracy of measurements? This is unclear and should be properly addressed.

9) If this study included a number of measurement campaigns, including various payloads or scenarios, all of them should be summarized in a table.

To sum up this paper requires major revisions before it can be processed further.

Author Response

We sincerely thank Reviewer 1 for the constructive comments and recommendations. The comments helped us improve the scientific framing, clarify the research gap, strengthen the related works, and better define the novelty and limitations of the study.

Comment 1:
In the Authors and Affiliation section check how to properly use Capital letters in case of your Institutions and make necessary corrections.

Response:
Thank you for this comment. We checked and corrected the capitalization and formatting of the institutional affiliations. The affiliation of the Center of Competence was revised and standardized. We also checked the correspondence line and removed formatting inconsistencies.

Changes in the manuscript:
The Authors and Affiliation section was revised. The institutional names were checked and corrected according to consistent capitalization and journal style.

Comment 2:
The technical specs of the drone, smartphone and camera should be summarized on the form of a table.

Response:
Thank you for this helpful suggestion. We added a dedicated table summarizing the main technical characteristics and functional role of the experimental components. The table includes the DJI Mini 4K UAV, the Servo King9000 smartphone, the UTi260M smartphone-connected thermal camera, the Ulefone Armor 27T smartphone, the UTi260T handheld thermal camera, the cord-based suspension, and the digital scale.

Changes in the manuscript:
A new table was added in the Materials and Methods section as Table 1: “Main technical characteristics and functional role of the experimental components.”

Comment 3:
There seems to be a change in font size/type from time to time. As an example, look at lines 233-237 on page 7. A simmilar change seams to appear in section 2.9 on page 11.

Response:
Thank you for pointing this out. We checked the manuscript formatting and revised inconsistent font size and font type issues. The formatting of the specified sections and tables was corrected according to the journal template.

Changes in the manuscript:
Font size, font type, table formatting, and section formatting were checked and corrected throughout the manuscript, especially in the Materials and Methods section and tables.

Comment 4:
Line 274 seems to be missing a beginning.

Response:
Thank you for identifying this issue. The incomplete sentence was corrected during the revision. The relevant methodological description was rewritten to improve clarity and continuity.

Changes in the manuscript:
The affected sentence in the Materials and Methods section was corrected and integrated into the revised methodological workflow description.

Comment 5:
The paper lacks a proper Related Works section, summing up previously published works on the aforementioned topic. Do summarize advancements in either drone technology, thermal inaging, sensors/cameras, etc.

Response:
We agree with the reviewer. We added and expanded a dedicated Related Works and Research Gap section. The revised section now discusses UAV thermal imaging, low-cost UAV thermal sensing, smartphone-connected thermal cameras, radiometric and non-radiometric thermal acquisition, UAV payload limitations, robust UAV operation under disturbances, and suspended-payload effects.

Changes in the manuscript:
Section 1.1, “Related Works and Research Gap,” was substantially expanded and reorganized. Additional references were added to support the revised discussion.

Comment 6:
The title includes the word smartphone, however very little attention is given to this mobile device.

Response:
Thank you for this important observation. We revised the manuscript to clarify the central role of the smartphone in the workflow. The Servo King9000 smartphone is now explicitly described as the central display and recording interface between the UTi260M thermal camera and the UAV RGB camera. We explain that without the smartphone, the UTi260M stream could not be operated, visualized, recorded locally, or indirectly captured by the UAV onboard RGB camera.

Changes in the manuscript:
A new subsection, “System Architecture and Role of the Smartphone,” was added in the Materials and Methods section. The role of the smartphone was also clarified in the Abstract, Highlights, Introduction, Results, and Discussion.

Comment 7:
Check the font size and type in case of tables and make necessary corrections according to the Journal template.

Response:
Thank you. We checked all tables and corrected their formatting according to the journal template. Table font size, alignment, and consistency were reviewed.

Changes in the manuscript:
All tables were checked and reformatted where necessary.

Comment 8:
After reading the whole paper I cannot notice the main aim of it. Was it the integration of a consumer drone + smartphone + camera? Was it the analysis of payload vs performance of the drone? Was it the stability or accuracy of measurements? This is unclear and should be properly addressed.

Response:
We agree that the original version did not state the aim clearly enough. The manuscript has been revised to define the main objective more explicitly. The objective is now stated as the experimental evaluation of the feasibility and operational limits of a minimum-cost indirect UAV–smartphone–thermal camera workflow under exploratory field conditions.

The revised manuscript clarifies that the study is not intended to validate a professional radiometric UAV thermal imaging platform. Instead, it evaluates whether a low-cost consumer UAV, lightweight smartphone, and smartphone-connected thermal camera can provide qualitative short-range hotspot presence/absence indication, and what operational limitations arise from payload mass, suspended-load oscillation, display readability, endurance reduction, post-flight motor heating, and lack of raw radiometric data.

Changes in the manuscript:
Section 1.2, “Objective and Contributions,” was rewritten. The Abstract, Introduction, Discussion, and Conclusions were also revised to reflect the clarified aim.

Comment 9:
If this study included a number of measurement campaigns, including various payloads or scenarios, all of them should be summarized in a table.

Response:
Thank you for the suggestion. We added a dedicated test matrix summarizing the experimental campaigns and measurement scenarios. The table includes no-payload baseline flight, payload hover flight, daylight display recording, nighttime display recording, altitude usability assessment, direct thermal recording, UAV-recorded display comparison, motor thermal inspection, and warning observation.

Changes in the manuscript:
A new table was added as Table 3: “Experimental campaigns and measurement scenarios used in the study.”

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript presents a pragmatic, hands-on evaluation of a cobbled-together thermal imaging system using a sub-250g consumer drone, a cheap smartphone, and a clip-on thermal camera. The core idea—using the drone’s own RGB camera to film the phone’s screen displaying the thermal feed—is clever in its simplicity and squarely aimed at budget-constrained users like educators or hobbyists. The main contribution is a clear, experimentally-derived map of the system's severe operational boundaries: a ~36.5% payload-to-UAV mass ratio that guts flight time by over 40%, induces asymmetric motor heating of nearly 20°C, and triggers "Propeller Error" warnings. The finding that nighttime operation dramatically improves the readability of the phone screen is intuitive but well-documented here, providing a practical guideline for anyone trying this.

Methodologically, the paper is thorough in its documentation of the physical penalties. Weighing everything, calculating moments, measuring motor temperatures with a second thermal camera—this is solid, reproducible engineering characterization. The use of image-based metrics like entropy and SNR to quantify the degradation from the indirect recording path is a reasonable approach given the lack of raw radiometric data.

My biggest issue is the fundamental premise of the "indirect" method. The authors acknowledge it, but I don't think they fully grapple with the consequences. Recording a phone screen with a drone camera introduces a cascade of uncontrollable degradations: auto-exposure changes on the drone camera will wash out the screen, the phone’s own auto-brightness fights you, rolling shutter artifacts from the drone camera interacting with the phone's screen refresh rate create bizarre banding, and the constant pendulum sway of the payload means the screen is never perfectly flat or in focus. The resulting video is not "thermal data"; it's a video of a video of thermal data. The paper would be much stronger if it explicitly stated that this workflow is only useful for presence/absence detection of a hot spot and is utterly useless for any kind of quantitative temperature measurement or even accurate shape delineation. The discussion dances around this but doesn't drive the knife in.

A specific weakness is the motor temperature analysis. Measuring the motor housing temperature with a handheld thermal camera after landing is fine for a rough comparison, but the cooling dynamics during the landing sequence are completely unaccounted for. A motor that was at 80°C in flight could cool to 42°C by the time the picture is taken 30 seconds later. The claim of a "~20°C increase" is thus a significant underestimate of the true thermal stress during flight. For a paper focused on operational limits, this is a critical blind spot. Real-time motor temperature logging via a telemetry protocol would have been far more informative.

Finally, the conclusions are appropriately cautious but miss a key opportunity for actionable advice. The suggestion for a "rigid lightweight mount" is obvious. A more expert insight would be to suggest a specific type of mount, like a carbon fiber rod that places the phone below the propeller wash entirely, or to recommend disabling the drone's downward-facing optical flow sensor, which likely went haywire trying to lock onto the swinging phone and contributed to the stability issues. The paper also fails to discuss the viability of simply using a longer USB cable to place the phone on the ground and only flying the tiny thermal camera—a common hack in the DIY drone community that avoids the entire payload mass problem. Ignoring this alternative makes the work feel less like an exploration of solutions and more like a report on a single, flawed implementation.

Author Response

We sincerely thank Reviewer 2 for the detailed technical comments. The comments helped us improve the engineering interpretation of the payload configuration, endurance reduction, motor thermal response, and operational limitations.

Comment 1:
This manuscript presents a pragmatic, hands-on evaluation of a cobbled-together thermal imaging system using a sub-250g consumer drone, a cheap smartphone, and a clip-on thermal camera. The core idea—using the drone’s own RGB camera to film the phone’s screen displaying the thermal feed—is clever in its simplicity and squarely aimed at budget-constrained users like educators or hobbyists. The main contribution is a clear, experimentally-derived map of the system's severe operational boundaries: a ~36.5% payload-to-UAV mass ratio that guts flight time by over 40%, induces asymmetric motor heating of nearly 20°C, and triggers "Propeller Error" warnings. The finding that nighttime operation dramatically improves the readability of the phone screen is intuitive but well-documented here, providing a practical guideline for anyone trying this. Methodologically, the paper is thorough in its documentation of the physical penalties. Weighing everything, calculating moments, measuring motor temperatures with a second thermal camera—this is solid, reproducible engineering characterization. The use of image-based metrics like entropy and SNR to quantify the degradation from the indirect recording path is a reasonable approach given the lack of raw radiometric data.

Response:
We sincerely thank the reviewer for this positive and constructive assessment. We are grateful that the reviewer recognized the practical engineering value of the proposed low-cost configuration and the documentation of its operational boundaries.

In the revised manuscript, we retained the practical focus of the study but clarified the scope more carefully. The system is now explicitly presented as a minimum-cost indirect UAV thermal sensing workflow and not as a professional radiometric UAV thermography platform. The physical penalties of the payload, including mass ratio, endurance reduction, post-flight motor temperature asymmetry, and operational warnings, are described as the main engineering findings.

Changes in the manuscript:
The Abstract, Highlights, Introduction, Results, Discussion, and Conclusions were revised to emphasize the practical engineering characterization and the operational limits of the tested configuration.

Comment 2:
My biggest issue is the fundamental premise of the "indirect" method. The authors acknowledge it, but I don't think they fully grapple with the consequences. Recording a phone screen with a drone camera introduces a cascade of uncontrollable degradations: auto-exposure changes on the drone camera will wash out the screen, the phone’s own auto-brightness fights you, rolling shutter artifacts from the drone camera interacting with the phone's screen refresh rate create bizarre banding, and the constant pendulum sway of the payload means the screen is never perfectly flat or in focus. The resulting video is not "thermal data"; it's a video of a video of thermal data. The paper would be much stronger if it explicitly stated that this workflow is only useful for presence/absence detection of a hot spot and is utterly useless for any kind of quantitative temperature measurement or even accurate shape delineation. The discussion dances around this but doesn't drive the knife in.

Response:
We fully agree with the reviewer. This was one of the most important points in the revision. We substantially strengthened the manuscript to explicitly state that the UAV-recorded output is not direct thermal data, but an RGB recording of a smartphone display showing thermal information.

The revised manuscript now clearly explains that the indirect display-recording workflow is affected by display brightness, ambient illumination, reflections, camera exposure, focus, display size in the frame, screen refresh effects, compression, and payload motion. We also explicitly state that the workflow is not suitable for quantitative temperature measurement, radiometric thermal mapping, or accurate thermal shape delineation.

The intended use was narrowed to short-range qualitative hotspot presence/absence indication under favorable low-glare or nighttime conditions.

Changes in the manuscript:
This clarification was added and strengthened in the Abstract, Highlights, Introduction, Related Works and Research Gap, Materials and Methods, Discussion, Limitations, and Conclusions. The manuscript now repeatedly emphasizes that the output is not radiometric thermal data.

Comment 3:
A specific weakness is the motor temperature analysis. Measuring the motor housing temperature with a handheld thermal camera after landing is fine for a rough comparison, but the cooling dynamics during the landing sequence are completely unaccounted for. A motor that was at 80°C in flight could cool to 42°C by the time the picture is taken 30 seconds later. The claim of a "~20°C increase" is thus a significant underestimate of the true thermal stress during flight. For a paper focused on operational limits, this is a critical blind spot. Real-time motor temperature logging via a telemetry protocol would have been far more informative.

Response:
We agree with the reviewer. The motor temperature measurements were obtained after landing using a handheld UTi260T thermal camera and therefore do not represent real-time in-flight motor temperatures. We revised the manuscript to make this limitation explicit.

The revised text now presents the measured values as post-flight motor-region temperatures and conservative indicators of payload-induced propulsion loading. We also state that the true in-flight motor temperature could have been higher due to cooling during landing and after shutdown. Real-time motor telemetry or embedded temperature logging is now identified as a recommended future improvement.

Changes in the manuscript:
The Results, Discussion, Limitations, and Conclusions were revised. The wording was changed from direct motor-temperature interpretation to post-flight motor-region thermal inspection. The limitation related to cooling dynamics and lack of real-time motor temperature logging was added.

Comment 4:
Finally, the conclusions are appropriately cautious but miss a key opportunity for actionable advice. The suggestion for a "rigid lightweight mount" is obvious. A more expert insight would be to suggest a specific type of mount, like a carbon fiber rod that places the phone below the propeller wash entirely, or to recommend disabling the drone's downward-facing optical flow sensor, which likely went haywire trying to lock onto the swinging phone and contributed to the stability issues. The paper also fails to discuss the viability of simply using a longer USB cable to place the phone on the ground and only flying the tiny thermal camera—a common hack in the DIY drone community that avoids the entire payload mass problem. Ignoring this alternative makes the work feel less like an exploration of solutions and more like a report on a single, flawed implementation.

Response:
Thank you for these useful suggestions. We revised the manuscript to discuss alternative configurations and practical improvements more explicitly. A direct rigid attachment and a lightweight composite or carbon-fiber support are now discussed as possible future solutions to reduce payload oscillation and improve the stability of the smartphone-displayed thermal stream.

We also added discussion of the direct-mounting alternative and the reason why the suspended display-facing configuration was used in this study: the purpose was to evaluate whether the UAV onboard RGB camera could record the smartphone display without modifying the UAV hardware or adding a dedicated transmission system.

The option of separating the smartphone and thermal camera through a longer USB cable is acknowledged conceptually as a possible alternative, but the present work focused on the tested configuration because the UTi260M operation depended on the smartphone interface and the goal was to assess the minimum-cost indirect display-recording workflow as implemented.

Regarding the downward optical flow sensor, we agree that the swinging payload may affect UAV stabilization. We revised the Discussion to treat payload oscillation, downwash interaction, center-of-gravity offset, and visual stabilization sensitivity as operational limitations. However, disabling manufacturer safety or stabilization functions was not tested in this study and therefore was not recommended as an experimentally validated procedure.

Changes in the manuscript:
A dedicated subsection on direct mounting and the reason for the suspended configuration was added. The Discussion and Future Work sections were expanded to include more specific recommendations for rigid lightweight or composite supports, reduced oscillation, improved alignment, and future configurations.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript presents an experimental evaluation of a low-cost indirect thermal sensing system mounted on a lightweight UAV. The proposed configuration combines a DJI Mini 4K platform, a smartphone, and a smartphone-connected thermal camera, where the UAV RGB camera records the smartphone-displayed thermal imagery during flight. The topic is timely and relevant. However, some concern should be addressed.

I consider that the scientific contribution remains unclear.

I consider that a weakness is the absence of repeated experiments and statistical validation.

Most conclusions are derived from a very limited number of observations.

No confidence intervals, standard deviations, or repeatability analyses are presented.

Section 2.8 introduces several image quality metrics including entropy, coefficient of variation, SNR, contrast, and uniformity index. However, these metrics are not meaningfully incorporated into the results section.

The manuscript develops a relatively extensive mathematical framework for image evaluation, but the corresponding quantitative analysis is largely absent.

I suggest to the authors present quantitative results for all introduced metrics.

Lines 278-280. The manuscript reports 0.0893 N, which is an order-of-magnitude error. This calculation should be corrected.

Lines 322-347. The authors acknowledge that radiometric data are unavailable. This limitation is important and should receive greater emphasis throughout the study.

Lines 472-488. The motor temperature analysis is based on representative observations rather than systematic measurements.

Lines 647-654. The statement regarding improved thermal contrast at night is reasonable, but no measurements are presented to support this claim within the study.

Coefficient of variation is incorrectly defined (Eq. 6)

Eq. (17) is incomplete.

The mathematical formulation contains several inconsistencies. The authors should carefully revise all equations and verify the correctness of the associated calculations.

The literature review (1° and 2° paragraphs) is primarily focused on thermal sensing applications and low-cost UAV platforms. In addition to sensing capabilities, practical UAV deployments often require robust flight performance under disturbances, payload variations, and environmental uncertainties, which has motivated extensive research on robust control strategies. The introduction could be strengthened by including representative studies addressing robust control (for example, robust backstepping-sliding control of a quadrotor UAV with disturbance compensation, Robust Nonlinear Control with Estimation of Disturbances and Parameter Uncertainties for UAVs and Integrated Brushless DC Motors)

Conclusions: several statements should be softened because the study is based on a limited number of experiments and lacks statistical validation.

Author Response

We sincerely thank Reviewer 3 for the critical and helpful comments. The comments were particularly useful in clarifying the non-radiometric nature of the workflow, the limited experimental scope, and the need to avoid unsupported statistical or detection-performance claims.

Comment 1:
The manuscript presents an experimental evaluation of a low-cost indirect thermal sensing system mounted on a lightweight UAV. The proposed configuration combines a DJI Mini 4K platform, a smartphone, and a smartphone-connected thermal camera, where the UAV RGB camera records the smartphone-displayed thermal imagery during flight. The topic is timely and relevant. However, some concern should be addressed.

Response:
We thank the reviewer for the constructive evaluation and for recognizing the relevance of the topic. We have revised the manuscript extensively to address the concerns raised. The revised manuscript now frames the system as an exploratory feasibility and operational-limits assessment of a low-cost indirect UAV thermal sensing workflow.

Changes in the manuscript:
The manuscript was revised throughout, especially in the Introduction, Materials and Methods, Results, Discussion, and Conclusions.

Comment 2:
I consider that the scientific contribution remains unclear.

Response:
We agree that the scientific contribution needed clearer definition. We revised the objective, research gap, and contribution sections. The manuscript now states that the novelty is not the development of a professional UAV thermal imaging system, but the experimental characterization of an intentionally minimum-cost indirect thermal sensing configuration operating close to the practical limits of a lightweight consumer UAV.

The contribution is now defined as the assessment of what can and cannot be achieved when a smartphone-connected thermal camera is integrated with a small UAV without a dedicated thermal payload, gimbal stabilization, raw radiometric export, or onboard thermal-video transmission.

Changes in the manuscript:
Section 1.1, “Related Works and Research Gap,” and Section 1.2, “Objective and Contributions,” were rewritten and expanded. The Abstract and Conclusions were also revised.

Comment 3:
I consider that a weakness is the absence of repeated experiments and statistical validation.

Response:
We agree with the reviewer. The revised manuscript now explicitly states that the study is exploratory and that the number of repeated trials was limited. We avoid presenting the results as statistically validated detection performance or radiometric validation.

The revised manuscript clarifies that the data should be interpreted as practical feasibility and operational-limit observations rather than statistically validated UAV thermal detection performance.

Changes in the manuscript:
The Materials and Methods, Results, Discussion, Limitations, and Conclusions were revised to clarify the exploratory nature of the study and the absence of statistical validation.

Comment 4:
Most conclusions are derived from a very limited number of observations.

Response:
We agree. The Conclusions were softened substantially. The revised Conclusions no longer make broad claims about general UAV thermal detection capability. Instead, they state that the proposed workflow is feasible only for short-range qualitative hotspot presence/absence indication under favorable low-glare or nighttime conditions.

Changes in the manuscript:
The Conclusions section was revised to limit the claims to the tested configuration and exploratory field observations.

Comment 5:
No confidence intervals, standard deviations, or repeatability analyses are presented.

Response:
Thank you for this comment. Because the study was exploratory and the number of repeated trials was limited, we did not add confidence intervals that would imply stronger statistical validity than the dataset supports. Instead, we revised the text to explicitly avoid statistical claims.

However, we added a short descriptive entropy analysis of consecutive smartphone-displayed thermal frames to document short-term frame-to-frame image-information stability within a selected region of interest. This analysis is clearly described as descriptive only and not as statistical validation, detection accuracy, or radiometric performance.

Changes in the manuscript:
A descriptive entropy analysis was added in the Methods and Results sections. The limitations regarding the absence of formal repeatability and confidence intervals were strengthened.

Comment 6:
Section 2.8 introduces several image quality metrics including entropy, coefficient of variation, SNR, contrast, and uniformity index. However, these metrics are not meaningfully incorporated into the results section.

Response:
We agree. To avoid introducing metrics that were not fully supported by the available data, we revised the manuscript and removed the unsupported extensive image-quality metric framework. The revised manuscript keeps only a short descriptive entropy analysis as a relative image-information indicator.

Changes in the manuscript:
The previous excessive image-quality metric framework was removed or reduced. A focused descriptive entropy analysis was added and discussed cautiously.

Comment 7:
The manuscript develops a relatively extensive mathematical framework for image evaluation, but the corresponding quantitative analysis is largely absent.

Response:
We agree. The mathematical framework was simplified to avoid unsupported quantitative claims. The revised manuscript now includes only the equations necessary for payload mass, payload ratio, payload force, destabilizing moment, endurance reduction, motor thermal response, motor thermal asymmetry, and descriptive usability interpretation.

Changes in the manuscript:
The mathematical formulation was revised and simplified. Unsupported or incomplete equations were removed.

Comment 8:
I suggest to the authors present quantitative results for all introduced metrics.

Response:
Thank you for the suggestion. Instead of retaining many image metrics without sufficient corresponding data, we revised the manuscript to avoid overextending the analysis. The unsupported metrics were removed, and only the descriptive entropy analysis was retained and presented quantitatively.

This approach was chosen to ensure that all introduced quantitative results are actually supported by the revised manuscript.

Changes in the manuscript:
The Results section now includes the descriptive entropy values for selected smartphone-displayed thermal frames. Other unsupported metrics were removed or not emphasized.

Comment 9:
Lines 278-280. The manuscript reports 0.0893 N, which is an order-of-magnitude error. This calculation should be corrected.

Response:
Thank you for identifying this calculation error. We corrected the payload force calculation. For an effective payload mass of 91 g, the payload weight force is calculated as:

F = 0.091 × 9.81 = 0.893 N.

Changes in the manuscript:
The payload force equation and numerical value were corrected in the payload suspension and camera-display geometry section.

Comment 10:
Lines 322-347. The authors acknowledge that radiometric data are unavailable. This limitation is important and should receive greater emphasis throughout the study.

Response:
We fully agree. The lack of raw radiometric data is now emphasized throughout the manuscript. We clearly state that the UAV-recorded video is not direct thermal data but an RGB recording of a smartphone display. The workflow is therefore not suitable for quantitative temperature measurement, radiometric thermal mapping, or accurate thermal shape delineation.

Changes in the manuscript:
This limitation was emphasized in the Abstract, Highlights, Introduction, Materials and Methods, Discussion, Limitations, and Conclusions.

Comment 11:
Lines 472-488. The motor temperature analysis is based on representative observations rather than systematic measurements.

Response:
We agree. The revised manuscript now clearly states that the motor temperature analysis is based on selected post-flight inspections and representative observations. The results are interpreted as conservative indicators of payload-induced propulsion loading rather than systematic real-time motor-temperature measurements.

Changes in the manuscript:
The Results and Discussion sections were revised. The limitations of post-flight motor inspection were explicitly added.

Comment 12:
Lines 647-654. The statement regarding improved thermal contrast at night is reasonable, but no measurements are presented to support this claim within the study.

Response:
We agree. The wording was revised to avoid unsupported claims of improved thermal contrast. The revised manuscript now states that nighttime and low-glare operation improved the readability of the smartphone-displayed thermal stream in the UAV RGB recording. This is presented as a qualitative usability observation, not as a quantitative thermal contrast measurement.

Changes in the manuscript:
The Results, Discussion, and Conclusions were revised to replace claims of improved thermal contrast with improved display readability and visual usability under low-glare or nighttime conditions.

Comment 13:
Coefficient of variation is incorrectly defined (Eq. 6)

Response:
Thank you for identifying this issue. The coefficient of variation equation was removed from the revised manuscript because the extended image-quality metric framework was simplified and not retained as a central quantitative analysis.

Changes in the manuscript:
The incorrect coefficient of variation equation was removed.

Comment 14:
Eq. (17) is incomplete.

Response:
Thank you for this comment. The incomplete equation was removed during the revision. The manuscript was checked to ensure that all remaining equations are complete and correctly numbered.

Changes in the manuscript:
The incomplete Eq. (17) was removed. Equation numbering was revised and checked.

Comment 15:
The mathematical formulation contains several inconsistencies. The authors should carefully revise all equations and verify the correctness of the associated calculations.

Response:
We agree. All equations and associated calculations were checked and corrected. The payload force calculation was corrected, unsupported image-quality equations were removed, and the remaining equations were aligned with the revised methodology.

Changes in the manuscript:
The equations and calculations were revised throughout the Materials and Methods and Results sections.

Comment 16:
The literature review (1° and 2° paragraphs) is primarily focused on thermal sensing applications and low-cost UAV platforms. In addition to sensing capabilities, practical UAV deployments often require robust flight performance under disturbances, payload variations, and environmental uncertainties, which has motivated extensive research on robust control strategies. The introduction could be strengthened by including representative studies addressing robust control (for example, robust backstepping-sliding control of a quadrotor UAV with disturbance compensation, Robust Nonlinear Control with Estimation of Disturbances and Parameter Uncertainties for UAVs and Integrated Brushless DC Motors)

Response:
Thank you for this valuable recommendation. We expanded the Related Works and Research Gap section to include UAV deployment under payload variation, wind disturbance, mechanical uncertainty, and robust control strategies. We added representative studies related to robust UAV control, disturbance compensation, nonlinear control, and integrated motor-control approaches.

Changes in the manuscript:
Section 1.1 was expanded to include robust UAV operation under disturbances and payload-induced uncertainty.

Comment 17:
Conclusions: several statements should be softened because the study is based on a limited number of experiments and lacks statistical validation.

Response:
We agree. The Conclusions were substantially softened. The revised Conclusions now state that the proposed workflow is feasible only as a short-range qualitative screening method and that it is not suitable for quantitative temperature measurement, radiometric thermal mapping, or accurate thermal shape delineation.

Changes in the manuscript:
The Conclusions section was revised to avoid broad claims and to emphasize the exploratory nature and limitations of the study.

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript presents a feasibility study of a low-cost indirect UAV thermal sensing workflow. In its current form, the work should be understood mainly as a preliminary screening approach, not as a substitute for professional radiometric UAV thermal platforms.

 

The introduction gives a reasonable general background to UAV thermal imaging and mentions several relevant application areas, including search and rescue, wildlife monitoring, agricultural inspection, building inspection, and low-cost sensing. However, the background remains too broad and does not sufficiently address the literature most directly connected to the proposed setup. A more focused discussion of smartphone-connected thermal cameras, indirect display-recording approaches, the payload limitations of sub-250 g UAVs, the distinction between radiometric and non-radiometric thermal data, and the safety or regulatory constraints associated with this type of flight configuration is expected. Some citations also seem only loosely connected to, or even mismatched with, the statements they are used to support. This weakens the credibility of the introduction. The literature review would benefit from a more selective and better justified reference base, since it currently relies rather heavily on recent MDPI-family publications and includes some engineering references that do not clearly support the UAV thermal sensing problem addressed in the manuscript.

 

The research design is acceptable for an exploratory feasibility study, especially because the authors are testing a low-cost configuration and do not appear to claim that it is a fully validated professional system. Even so, the experimental basis is too limited for broad conclusions. The study uses a single UAV, a single payload configuration, a small number of trials, and only partially defined environmental conditions. Important factors such as wind, ambient temperature, illumination, battery state, flight profile, target properties, and the exact operating conditions should be reported more clearly. Without this information, it is difficult to judge whether the observed limitations are due to the proposed setup itself or to the specific test conditions.

 

The methods section of the manuscript contains useful and practical information on the hardware, payload mass, suspension arrangement, indirect acquisition logic, endurance calculation, and motor-temperature comparison. However, several methodological details are still missing. The manuscript should include a clear test matrix; the number of flights performed under each condition; calibration and emissivity settings; battery start and stop rules; target dimensions and distances; image-processing workflow; software settings; and how uncertainty was treated. Some equations and metrics also require correction before publication:

1. The payload weight calculation: 0.091 kg x 9.81 m/s² is 0.893 N, not 0.0893 N.
2. The coefficient of variation formula is inverted
3. The Spearman correlation is not properly defined,
4. Several formulas contain formatting or typographical problems.

 

The results are generally understandable, and the main tendencies are clear. The proposed payload can be lifted; endurance decreases substantially; motor temperatures increase; display readability is better at night; and the practical altitude range is limited. These are useful findings for a feasibility-oriented paper. The presentation would nevertheless be stronger if measured data were more clearly separated from interpretation, if uncertainty ranges were added, and if all trials were reported rather than only representative examples. The figures also need further attention, both in terms of quality and consistency.

 

Table 8 needs revision - one motor measurement is cooler with the payload than without it, but the surrounding text presents the loaded configuration more generally as producing higher motor temperatures. Either this should be explained, or the wording should be made more precise. Figure captions and numbering should also be checked carefully. This is particularly important for the exploratory relationship figure, where the caption does not fully correspond to all displayed subplots.

 

When considering the manuscript as a small feasibility demonstration, the claims about short-duration screening, endurance loss, motor loading, and the limitations of indirect non-radiometric display capture are mostly supported in the conclusions. However, they are not sufficiently supported for broader claims about target detectability or recommended operating windows. The authors should expand the manuscript with ground-truth targets, repeated measurements, controlled conditions, and quantitative detection criteria.

 

The English is generally OK, but the manuscript still needs revision for language and style. Several sentences read like draft text rather than a polished scientific article. There are duplicated words, awkward phrases, inconsistent terminology, broken equations, and some inappropriate wording. For instance, describing the work as retrospective does not seem suitable, since the study appears to be a prospective experimental test.

 

The reference list needs revision. The self-related or local-group references need better justification or should be simply removed unless they directly support the specific method or interpretation. Many references are relevant to UAV thermal imaging in a broad sense, but the numbering does not follow the claims made in the introduction. Several references appear unrelated to the topic, two placeholder references remain, and some entries should be checked for year, title, DOI, and citation purpose.

 

Suggestion for authors:

 

1. The authors should rebuild the introduction and reference list so that each citation directly supports the statement to which it is attached. It should particularly cover low-cost thermal modules, smartphone thermography, UAV payload effects, indirect optical capture, and the limitations of non-radiometric thermal data.
2. The experimental design and methods should be strengthened by adding a clear test matrix, repeated flights, controlled environmental conditions, defined targets, calibration settings, raw data handling, uncertainty estimates, and corrected mathematical expressions.
3. The results should be reorganised around measured evidence; complete datasets should be included rather than only representative examples; the figures should be improved; and repeated explanatory text should be reduced.

Author Response

We sincerely thank Reviewer 4 for the constructive comments and suggestions. The comments helped us improve the structure, formatting, technical consistency, and clarity of the revised manuscript.

Comment 1:
The manuscript presents a feasibility study of a low-cost indirect UAV thermal sensing workflow. In its current form, the work should be understood mainly as a preliminary screening approach, not as a substitute for professional radiometric UAV thermal platforms.

Response:
We fully agree. The revised manuscript now explicitly presents the work as an exploratory feasibility and operational-limits assessment of a minimum-cost indirect UAV thermal sensing workflow. We clearly state that the system is not a substitute for professional radiometric UAV thermal platforms.

Changes in the manuscript:
The Abstract, Introduction, Related Works and Research Gap, Materials and Methods, Discussion, Limitations, and Conclusions were revised to emphasize the preliminary and non-radiometric nature of the workflow.

Comment 2:
The introduction gives a reasonable general background to UAV thermal imaging and mentions several relevant application areas, including search and rescue, wildlife monitoring, agricultural inspection, building inspection, and low-cost sensing. However, the background remains too broad and does not sufficiently address the literature most directly connected to the proposed setup. A more focused discussion of smartphone-connected thermal cameras, indirect display-recording approaches, the payload limitations of sub-250 g UAVs, the distinction between radiometric and non-radiometric thermal data, and the safety or regulatory constraints associated with this type of flight configuration is expected. Some citations also seem only loosely connected to, or even mismatched with, the statements they are used to support. This weakens the credibility of the introduction. The literature review would benefit from a more selective and better justified reference base, since it currently relies rather heavily on recent MDPI-family publications and includes some engineering references that do not clearly support the UAV thermal sensing problem addressed in the manuscript.

Response:
Thank you for this detailed comment. We revised the Introduction and Related Works sections to make the literature review more focused and better connected to the proposed setup. The revised version now discusses smartphone-connected thermal cameras, low-cost thermal sensing, indirect display-recording limitations, UAV payload constraints, radiometric versus non-radiometric thermal data, and suspended-payload effects.

We also checked the reference list and revised the use of citations so that the references are better aligned with the statements they support.

Changes in the manuscript:
The Introduction and Section 1.1 were substantially revised. The reference base was checked and adjusted for better relevance.

Comment 3:
The research design is acceptable for an exploratory feasibility study, especially because the authors are testing a low-cost configuration and do not appear to claim that it is a fully validated professional system. Even so, the experimental basis is too limited for broad conclusions. The study uses a single UAV, a single payload configuration, a small number of trials, and only partially defined environmental conditions. Important factors such as wind, ambient temperature, illumination, battery state, flight profile, target properties, and the exact operating conditions should be reported more clearly. Without this information, it is difficult to judge whether the observed limitations are due to the proposed setup itself or to the specific test conditions.

Response:
We agree. The revised manuscript now explicitly frames the study as exploratory and avoids broad conclusions. We added a test matrix summarizing the experimental campaigns and measurement scenarios. We also clarified that environmental conditions such as wind, illumination, ambient temperature, and battery state were not controlled using laboratory instrumentation and should be treated as practical field variables and limitations.

Changes in the manuscript:
A test matrix was added as Table 3. The Materials and Methods and Limitations sections were revised to clarify the experimental scope and field-condition limitations.

Comment 4:
The methods section of the manuscript contains useful and practical information on the hardware, payload mass, suspension arrangement, indirect acquisition logic, endurance calculation, and motor-temperature comparison. However, several methodological details are still missing. The manuscript should include a clear test matrix; the number of flights performed under each condition; calibration and emissivity settings; battery start and stop rules; target dimensions and distances; image-processing workflow; software settings; and how uncertainty was treated.

Response:
Thank you for this comment. We added a clear test matrix to summarize the measurement scenarios and observations. The revised manuscript also clarifies the exploratory nature of the field tests and states that environmental conditions and uncertainties were not controlled using laboratory instrumentation.

The image-processing workflow was also clarified in the qualitative usability and entropy sections. We described that selected smartphone-displayed frames were converted to 8-bit grayscale and analyzed using identical regions of interest, while interface elements and overlays were excluded as far as possible.

Because the workflow is non-radiometric and raw radiometric data were not available, calibration and emissivity settings were not used for UAV-recorded display frames. This limitation is now explicitly stated.

Changes in the manuscript:
The Materials and Methods section was revised. A test matrix was added as Table 3. The qualitative image-usability and entropy analysis descriptions were clarified. The lack of raw radiometric data and the resulting calibration limitations were emphasized.

Comment 5:
Some equations and metrics also require correction before publication:

1. The payload weight calculation: 0.091 kg x 9.81 m/s² is 0.893 N, not 0.0893 N.
2. The coefficient of variation formula is inverted
3. The Spearman correlation is not properly defined,
4. Several formulas contain formatting or typographical problems.

Response:
We agree and thank the reviewer for identifying these issues. The payload weight calculation was corrected to 0.893 N. The coefficient of variation and Spearman correlation equations were removed because the extended image-quality metric framework was simplified. All remaining equations were checked, corrected, and renumbered.

Changes in the manuscript:
The payload force calculation was corrected. The unsupported or problematic equations were removed. All remaining equations were checked for consistency and formatting.

Comment 6:
The results are generally understandable, and the main tendencies are clear. The proposed payload can be lifted; endurance decreases substantially; motor temperatures increase; display readability is better at night; and the practical altitude range is limited. These are useful findings for a feasibility-oriented paper. The presentation would nevertheless be stronger if measured data were more clearly separated from interpretation, if uncertainty ranges were added, and if all trials were reported rather than only representative examples. The figures also need further attention, both in terms of quality and consistency.

Response:
Thank you for this constructive comment. We revised the Results section to separate measured observations from interpretation more clearly. The manuscript now presents payload mass, endurance reduction, post-flight motor temperature observations, display readability, and operating height limitations in a more cautious way.

We also revised figure captions, table captions, and figure/table citations to improve consistency. Because the study was exploratory and included representative field observations rather than a statistically controlled experimental campaign, we avoided adding uncertainty ranges that would imply unsupported statistical validity.

Changes in the manuscript:
The Results, Discussion, figure captions, table captions, and limitations were revised. Figures and tables were checked for consistency.

Comment 7:
Table 8 needs revision - one motor measurement is cooler with the payload than without it, but the surrounding text presents the loaded configuration more generally as producing higher motor temperatures. Either this should be explained, or the wording should be made more precise. Figure captions and numbering should also be checked carefully. This is particularly important for the exploratory relationship figure, where the caption does not fully correspond to all displayed subplots.

Response:
Thank you for identifying this issue. We revised the motor-temperature interpretation. The manuscript now avoids stating that every motor was hotter under payload conditions. Instead, it states that the maximum observed post-flight motor-region temperature and thermal asymmetry increased under payload operation.

We also checked and corrected figure captions, table captions, figure numbering, and table numbering. The final summary table was renumbered to avoid duplication.

Changes in the manuscript:
The Results and Discussion text around motor temperature was revised. Table numbering and figure captions were checked and corrected.

Comment 8:
When considering the manuscript as a small feasibility demonstration, the claims about short-duration screening, endurance loss, motor loading, and the limitations of indirect non-radiometric display capture are mostly supported in the conclusions. However, they are not sufficiently supported for broader claims about target detectability or recommended operating windows. The authors should expand the manuscript with ground-truth targets, repeated measurements, controlled conditions, and quantitative detection criteria.

Response:
We agree that broader claims about target detectability and recommended operating windows would require ground-truth targets, repeated measurements, controlled conditions, and quantitative detection criteria. Since such a dataset was not available in the present exploratory study, we did not expand the claims beyond the supported level.

Instead, we revised the manuscript to clearly limit the operating height range to an observed display-readability window for the tested configuration. We also state that the workflow is suitable only for qualitative hotspot presence/absence indication and not for validated detection performance.

Changes in the manuscript:
The Discussion and Conclusions were revised to avoid unsupported general detection claims and to emphasize the need for controlled repeated studies in future work.

Comment 9:
The English is generally OK, but the manuscript still needs revision for language and style. Several sentences read like draft text rather than a polished scientific article. There are duplicated words, awkward phrases, inconsistent terminology, broken equations, and some inappropriate wording. For instance, describing the work as retrospective does not seem suitable, since the study appears to be a prospective experimental test.

Response:
Thank you for this observation. We revised the language and style throughout the manuscript. Awkward phrases, inconsistent terminology, duplicated words, and problematic wording were corrected. The study is now described as an experimental feasibility and operational-limits assessment rather than as retrospective.

Changes in the manuscript:
The manuscript was revised for language, style, terminology, and formatting throughout.

Comment 10:
The reference list needs revision. The self-related or local-group references need better justification or should be simply removed unless they directly support the specific method or interpretation. Many references are relevant to UAV thermal imaging in a broad sense, but the numbering does not follow the claims made in the introduction. Several references appear unrelated to the topic, two placeholder references remain, and some entries should be checked for year, title, DOI, and citation purpose.

Response:
We agree. The reference list was checked and revised. References were reviewed for relevance to the revised claims, especially in the Introduction and Related Works sections. The manuscript now uses a more focused reference base related to UAV thermal imaging, low-cost thermal sensing, smartphone-connected thermal cameras, payload limitations, radiometric and non-radiometric data, and suspended-load operation.

Changes in the manuscript:
The reference list and in-text citation use were checked and revised.

Comment 11:
Suggestion for authors:

1. The authors should rebuild the introduction and reference list so that each citation directly supports the statement to which it is attached. It should particularly cover low-cost thermal modules, smartphone thermography, UAV payload effects, indirect optical capture, and the limitations of non-radiometric thermal data.

Response:
Thank you for the suggestion. We rebuilt and refocused the Introduction and Related Works sections. The revised text now better covers low-cost thermal modules, smartphone-connected thermography, UAV payload effects, indirect display-based optical capture, and limitations of non-radiometric thermal data.

Changes in the manuscript:
The Introduction, Related Works and Research Gap, and reference list were revised.

Comment 12:
Suggestion for authors:
2. The experimental design and methods should be strengthened by adding a clear test matrix, repeated flights, controlled environmental conditions, defined targets, calibration settings, raw data handling, uncertainty estimates, and corrected mathematical expressions.

Response:
Thank you. We added a clear test matrix and corrected the mathematical expressions. We also clarified the experimental limitations, including limited repeated trials, uncontrolled field variables, lack of raw radiometric data, and the qualitative nature of the workflow.

Because the present work is an exploratory feasibility study, we did not add unsupported uncertainty estimates or claim controlled experimental validation. Instead, we explicitly state that controlled repeated experiments, defined ground-truth targets, calibration procedures, and quantitative uncertainty analysis should be addressed in future work.

Changes in the manuscript:
Table 3 was added. The Materials and Methods, equations, Results, Discussion, and Limitations were revised.

Comment 13:
Suggestion for authors:
3. The results should be reorganised around measured evidence; complete datasets should be included rather than only representative examples; the figures should be improved; and repeated explanatory text should be reduced.

Response:
We revised the Results section to better organize the findings around measured and observed evidence. We also improved figure and table captions, corrected numbering, added a descriptive entropy analysis, and reduced unsupported or repetitive explanatory text.

Changes in the manuscript:
The Results, Discussion, figure captions, table captions, and final summary table were revised.

Final note to the Academic Editor and Reviewers

We again thank the Academic Editor and all reviewers for their constructive comments. The revised manuscript has been substantially improved in terms of scientific framing, methodological clarity, technical consistency, and interpretation of limitations. We believe that the revised version now presents a clearer and more cautious feasibility assessment of a minimum-cost indirect UAV thermal sensing workflow.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Now the changes made are properly justified, I am also satisfied with the quality of responses based on reviewer comments. All modifications are clear and legible.
Taking all into consideration, I do recommend the paper to be processed further. It seems to fulfil necessary requirements in order to be accepted and published.

Author Response

Reviewer 1 comment:
Now the changes made are properly justified, I am also satisfied with the quality of responses based on reviewer comments. All modifications are clear and legible. Taking all into consideration, I do recommend the paper to be processed further. It seems to fulfil necessary requirements in order to be accepted and published.

Response:
We sincerely thank the reviewer for the positive assessment of the revised manuscript and for recognizing that the previous concerns were properly addressed. We appreciate the reviewer’s recommendation for further processing of the paper.

In the current revision, we made additional minor but important improvements to further strengthen the manuscript. These include correction of citation matching, clarification of motor-region thermal interpretation, refinement of terminology, and removal of potentially ambiguous ethical wording. We believe that these changes further improve the clarity and reliability of the manuscript.

Reviewer 2 Report

Comments and Suggestions for Authors

Accept.

Author Response

Reviewer 2 comment:
Accept.

Response:
We sincerely thank the reviewer for the positive recommendation and for supporting acceptance of the manuscript.

Although no further changes were requested by the reviewer, we revised the manuscript further in response to the remaining comments from another reviewer. These changes do not alter the main conclusions of the study, but improve citation accuracy, terminology, methodological transparency, and the explanation of limitations.

Reviewer 3 Report

Comments and Suggestions for Authors

I have reviewed the updated response and am satisfied that the authors have adequately addressed the previous concerns. I have no further comments and recommend acceptance of the manuscript in its present form.

Author Response

Reviewer 3 comment:
I have reviewed the updated response and am satisfied that the authors have adequately addressed the previous concerns. I have no further comments and recommend acceptance of the manuscript in its present form.

Response:
We sincerely thank the reviewer for the careful reassessment of the manuscript and for recommending acceptance. We appreciate the reviewer’s confirmation that the previous concerns were adequately addressed.

In the current revision, we made additional targeted improvements in response to the remaining comments from another reviewer. These changes were introduced to further improve the technical precision and editorial consistency of the manuscript.

Reviewer 4 Report

Comments and Suggestions for Authors

Dear Authors, thank you for sending the updated version of your manuscript. After careful proofreading, I can state that you have responded substantially to the review by reframing the study as a minimum-cost indirect UAV thermal sensing workflow rather than as a substitute for professional radiometric UAV thermography.

The revised manuscript now states more clearly that the UAV-recorded output is only an RGB recording of a smartphone display showing thermal information, and that the system is suitable only for short-range qualitative hotspot presence/absence indication under favourable low-glare or night-time conditions.

The introduction, research gap and discussion of operational limitations have been strengthened, and the major calculation error in the payload force has been corrected.

However, several important issues remain only partly addressed: the experimental basis is still limited to one UAV, one payload configuration and largely uncontrolled field conditions; ground-truth targets, repeated measurements, calibration details and uncertainty treatment are still insufficient; the motor-temperature anomaly in Table 8 is not fully explained; and some citation, terminology and language problems persist, including questionable reference matching and the inappropriate use of “retrospective” for an experimental study.

Overall, version 2 is clearly improved and responds to many of the main concerns, but further revision is still required before publication.

Author Response

Reviewer 4 comment:
Dear Authors, thank you for sending the updated version of your manuscript. After careful proofreading, I can state that you have responded substantially to the review by reframing the study as a minimum-cost indirect UAV thermal sensing workflow rather than as a substitute for professional radiometric UAV thermography.

The revised manuscript now states more clearly that the UAV-recorded output is only an RGB recording of a smartphone display showing thermal information, and that the system is suitable only for short-range qualitative hotspot presence/absence indication under favourable low-glare or nighttime conditions.

The introduction, research gap and discussion of operational limitations have been strengthened, and the major calculation error in the payload force has been corrected.

However, several important issues remain only partly addressed: the experimental basis is still limited to one UAV, one payload configuration and largely uncontrolled field conditions; ground-truth targets, repeated measurements, calibration details and uncertainty treatment are still insufficient; the motor-temperature anomaly in Table 8 is not fully explained; and some citation, terminology and language problems persist, including questionable reference matching and the inappropriate use of “retrospective” for an experimental study.

Overall, version 2 is clearly improved and responds to many of the main concerns, but further revision is still required before publication.

Response:
We sincerely thank the reviewer for the detailed and constructive assessment of the revised manuscript. We appreciate the reviewer’s recognition that the study has been substantially reframed as a minimum-cost indirect UAV thermal sensing workflow rather than as a substitute for professional radiometric UAV thermography.

We also thank the reviewer for identifying the remaining issues. These comments were highly useful, and we have addressed them carefully in the current revision. The manuscript has been revised in the following ways.