Integration of Payload Sensors to Enhance UAV-Based Spraying

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript describes the study of a spraying system mounted on a flying drone. The test set consists of a camera detecting "bad" plants and a sprayer model.

(1) In line 10 the authors talk about the system's resilience. However, if the article is recognized by readers only by Introduction, it may be incomprehensible.

(2) The authors presented a short analysis of the research available in the literature. However, they did not address the issues that are the purpose of their research described in this article. This research is interesting and it would be worth finding out if anyone has done anything similar or how the authors came up with the idea for such a solution.

(3) Unfortunately, the authors did not address the problem of application of the preparation by flying drones. The study conducted does not confirm the correctness of the application. A review of the literature in this area would be helpful.

(4) The test site and research methodology are sufficiently described.

(5) The drawings are clear and clearly visible.

(6) The conclusions should also include the most important results from the conducted test.

 

Author Response

The manuscript describes the study of a spraying system mounted on a flying drone. The test set consists of a camera detecting "bad" plants and a sprayer model.

Dear reviewer,

Thank you very much for your valuable comments. Please be assured that we have addressed them, as we believe they will significantly improve our manuscript in the next version. Please note that the changes and responses to your comments are highlighted in Emerald (light green) in the revised manuscript.

 

(1) In line 10 the authors talk about the system's resilience. However, if the article is recognized by readers only by Introduction, it may be incomprehensible.

In order to clarify the explanations about the system's resilience and stability, we modified the abstract conclusion, adding the following text.

Additionally, the UAV's flight controller demonstrated robust performance, maintaining stability despite the challenges posed by liquid load oscillations and varying payloads during the spraying process.

(2) The authors presented a short analysis of the research available in the literature. However, they did not address the issues that are the purpose of their research described in this article. This research is interesting and it would be worth finding out if anyone has done anything similar or how the authors came up with the idea for such a solution.

To address this suggestion, we have included in the introduction a discussion on the existing commercially available technologies. From there, we highlight that our approach aims to provide a similar solution but from an academic perspective, where the integration of sensors not only aids in UAV flight stabilization but also supports the execution of specific missions, such as pesticide application. The main challenge and purpose of this work is to demonstrate how sensor feedback can contribute to the overall task execution. In other words, we explore how load sensor data can be utilized as input for the state machine responsible for the mission of monitoring, detection, and pest control in crops.

Please find below the added text, as well as the references cited.

The approach to agricultural spraying using drones can be significantly enhanced by considering the weight of the pesticide being transported, rather than relying solely on the nozzle's spraying rate, as is common in commercial models, such as those discussed by \cite{mogili2018review} and DJI's Agras series\footnote{\url{https://ag.dji.com/}}.

 

 

(3) Unfortunately, the authors did not address the problem of application of the preparation by flying drones. The study conducted does not confirm the correctness of the application. A review of the literature in this area would be helpful.

The purpose of this work is to provide insight into the state of the art in agricultural pesticide application using UAVs, where the liquid payload weight is monitored throughout the flight. In line with your previous suggestion, we have included information that aligns our work with the development of flight strategies considering parametric variations, specifically the weight of the liquid payload. Regarding the literature review, we believe it is more beneficial to present what is currently available in the market from a commercial standpoint, thus motivating researchers to explore this field further.

An additional point of the work is to demonstrate that our approach can be extended to non-liquid payloads, such as using drones for seed distribution in reforestation missions or deploying larvicides in water reservoirs to combat dengue and similar diseases.

(4) The test site and research methodology are sufficiently described.

Thanks for your comment.

(5) The drawings are clear and clearly visible.

Thanks for your comment.

(6) The conclusions should also include the most important results from the conducted test.

In response to your suggestion, we would like to include the following text in the conclusion of the reply to your third comment.

Additionally, our approach can be extended to non-liquid payloads, as the load sensor operates independently of the type of cargo being transported. This enables the use of drones for seed distribution in reforestation missions or for deploying larvicides in water reservoirs to combat dengue and similar diseases.

Reviewer 2 Report

Comments and Suggestions for Authors

According to the summary, this work focuses on the use of load sensors to help with spraying tasks using unmanned 1 aerial vehicles (UAVs). To simulate the application of agricultural pesticides, the UAV follows 3 a predefined route and a computer vision system detects the presence of diseased plants, while the UAV after 4 detection, pauses its route momentarily and activates the spraying device.

Also, the paper foresees that when the storage tank is empty or the remaining quantity is insufficient for 7 another operation, the system commands the UAV to return to the base station for refueling, with 8 Experimental validations carried out in an indoor controlled environment.

As a result, the authors declare proved that the algorithm and hardware (sensors, controllers …) are robust despite variations in the payload and in the center of buoyancy during flight. In this particular fact, I do agree.

I congratulate the authors for such an interesting, relevant and well performed research. I have no major comments, and to my extent it could be published straight away.

However, in my opinion the title promises a too broad project, that could be interpreted in many ways. For example, according to the title, I would have expected the use of the load cells, not only for refilling purposes, but for some type of verification of the instant flow, while correcting the influence of the UAV orientation on the measuring system.

Reflecting on this opinion, I have concluded that, being these details important issues to me, they are most probably beyond the scope of the paper. Therefore, I recommend the improvement of the title by stating a more specific task more in the line of the actual objectives and results.

PS. The mathematical explanation is very clear and acppreciated.

Author Response

According to the summary, this work focuses on the use of load sensors to help with spraying tasks using unmanned aerial vehicles (UAVs). To simulate the application of agricultural pesticides, the UAV follows a predefined route and a computer vision system detects the presence of diseased plants, while the UAV after detection, pauses its route momentarily and activates the spraying device.

Also, the paper foresees that when the storage tank is empty or the remaining quantity is insufficient for another operation, the system commands the UAV to return to the base station for refueling, with Experimental validations carried out in an indoor controlled environment.

As a result, the authors declare proved that the algorithm and hardware (sensors, controllers …) are robust despite variations in the payload and in the center of buoyancy during flight. In this particular fact, I do agree.

I congratulate the authors for such an interesting, relevant and well performed research. I have no major comments, and to my extent it could be published straight away.

Dear reviewer,

Thank you for your thoughtful comments on our work and for signaling its acceptance. We hope that this revised version will meet the expectations of the other reviewers, allowing them to appreciate how our research can contribute to the academic community exploring this topic.

In response to your suggestions, we have highlighted your comments in Mulberry color in the revised text to make it easier for you to identify where we have incorporated your recommendations.

 

However, in my opinion the title promises a too broad project, that could be interpreted in many ways. For example, according to the title, I would have expected the use of the load cells, not only for refilling purposes, but for some type of verification of the instant flow, while correcting the influence of the UAV orientation on the measuring system.

Reflecting on this opinion, I have concluded that, being these details important issues to me, they are most probably beyond the scope of the paper. Therefore, I recommend the improvement of the title by stating a more specific task more in the line of the actual objectives and results.

 

Indeed, the proposal to use load sensors extends beyond just the use of liquid payloads for pesticide application tasks and determining when to return to base for refueling. In light of this and in direct alignment with your comment, we have added the following paragraph in the results and discussion section.

It is worthy to highlight that our approach can be adapted for non-liquid payloads since the load sensor functions independently of the cargo type. This versatility allows for the utilization of drones in various applications, such as seed distribution in reforestation efforts or the deployment of larvicides in water reservoirs to combat dengue and other similar diseases. This potential for diverse applications highlights the broader implications of our work in advancing UAV technology for environmental and public health initiatives.

 

PS. The mathematical explanation is very clear and appreciated.

Thanks for your comment.

Reviewer 3 Report

Comments and Suggestions for Authors

The author proposed a method to integrate effective payload sensors to enhance UAV spraying, associating the UAV payload with distribution to improve spraying efficiency. Additionally, the author developed a plant health monitoring system. Before publication, I suggest the following revisions:

1. The author introduced a UAV pesticide-targeted spraying system. How does this research differ from current commercial agricultural UAVs?
2. The author described their UAV system and provided experimental videos, but the paper lacks sufficient experimental results to support the findings.
3. The author should succinctly summarize the contributions of the paper in both the abstract and the conclusion.
4. I recommend that the author conduct additional experiments and compare the results with current research.
5. In Figure 4, "Good Plant," "Sick Plant," and "No Plant" are distinguished solely by color. This does not align with lines 71-72 in the text, which state, “We have also developed a drone-based plant health monitoring system that utilizes image processing to detect diseases and nutritional deficiencies based on plant coloration.” This is merely a simplified experimental substitute. If possible, the author should test on a real plant disease dataset.
6. In Figure 5, the author shows changes in load weight. Are these results derived directly from pump switch data? Are such predictions accurate? Has the author compared them with ground truth?
7. In Figure 6, the author outlines the overall experimental process, including UAV flight paths and corresponding plant disease classifications. I suggest moving this figure earlier in the paper, before the experimental results section, to provide readers with a clearer understanding of the experiment's process. Additionally, the author should clarify in both the figure and the text the advantages of this method and how it differs from current research.

Author Response

The author proposed a method to integrate effective payload sensors to enhance UAV spraying, associating the UAV payload with distribution to improve spraying efficiency. Additionally, the author developed a plant health monitoring system.

Dear reviewer,

Thank you for the time you dedicated to reviewing our work. I hope to clarify your doubts through our responses to your questions. To facilitate the visualization of the corrected or added sections in the text based on your suggestions, we have highlighted your comments in RoyalBlue in the updated version of the manuscript.

 

Before publication, I suggest the following revisions:

The author introduced a UAV pesticide-targeted spraying system. How does this research differ from current commercial agricultural UAVs?

Excellent question. Typically, commercial drones employ a strategy similar to the one used in our work to measure the load in the transport compartment. However, the information obtained is primarily used to determine whether the UAV should return to base for refueling. To the best of our knowledge, sensory readings are not utilized during spraying to ensure that identical amounts of inputs are applied to the crop. Our literature review indicates that commercial UAVs apply inputs at a constant rate based on the type of crop and/or navigation route. In this regard, it is important to highlight that our proposal considers the information regarding the amount of liquid in the tank, both to monitor the quantity of inputs applied and to decide whether it is feasible to apply the predetermined amount to a plant detected as diseased.

To emphasize the commercial aspect of UAVs in comparison to our proposal, the following text has been added to the new version of the paper:

The approach to agricultural spraying using drones can be significantly enhanced by considering the weight of the pesticide being transported, rather than relying solely on the nozzle's spraying rate, as is common in commercial models, such as those discussed by \cite{mogili2018review} and DJI's Agras series\footnote{\url{https://ag.dji.com/}}. Monitoring the weight provides a direct and accurate measurement of the actual amount of pesticide available, allowing for dynamic adjustments in the spraying process to optimize liquid distribution, minimize waste, and ensure uniform coverage. In contrast, systems that use the nozzle's spraying rate base their calculations on fixed parameters, which may not adequately reflect real variations in the field, such as changes in liquid viscosity or the performance of the pumping system due to wear on the nozzles over time. Thus, the weight-based approach offers greater accuracy and efficiency, particularly in variable field conditions where real-time adjustments are essential for effective and sustainable spraying.

 

The author described their UAV system and provided experimental videos, but the paper lacks sufficient experimental results to support the findings.

We appreciate your consideration, and to address this, we are providing several videos of the various experiments conducted during the experimental validation. In these videos, we aim to validate the refueling moment, the navigation speed along the reference route, the flow/quantity of liquid expelled after detecting a diseased plant, the position of the dispersion nozzle in relation to the onboard camera and the plant on the ground, among other experimental configurations. After conducting all these experimental validations, we concluded that the results that best represented and justified the proposal of the work were those shown in the video included in the original version of the paper (see link https://youtu.be/lB\_JT8X3N-8). 

With that said, I would like to emphasize that we performed a series of experimental tests and filtered those that provided the best academic exemplification.

We include this text at the end of Results and Discussion section:

In conclusion, we conducted additional tests to validate alternative approaches, which are showcased in the accompanying video (https://youtu.be/IdVl9vTMFpg). Throughout our experiments, we explored various configurations, including adjustments to the control strategy, the simulated liquid dispersion system, and the monitoring and pesticide application strategies. These findings underscore the effectiveness of our proposed methods and provide further insight into the potential of UAV-based spraying systems.

 

The author should succinctly summarize the contributions of the paper in both the abstract and the conclusion.

Thank you very much for the suggestion. To address it, we have added the following phrase in the abstract and the conclusion:

In summary, our main contribution is a real-time payload monitoring system that continuously tracks weight during flight, ensuring accurate pesticide application by preventing over- or under-spraying. The system also aids in automatic refueling by detecting low pesticide levels and signaling the UAV to return to base when necessary.

 

I recommend that the author conduct additional experiments and compare the results with current research.

As mentioned above, we conducted several experiments with various configurations, including adjustments to the control strategy, simulated liquid dispersion system, and monitoring and pesticide application strategy. We concluded that the results presented in the paper best demonstrate our proposal from an academic standpoint. Therefore, we believe that comparisons with other results are beyond the scope of this work, whose contribution was addressed in the previous item.

 

In Figure 4, "Good Plant," "Sick Plant," and "No Plant" are distinguished solely by color. This does not align with lines 71-72 in the text, which state, “We have also developed a drone-based plant health monitoring system that utilizes image processing to detect diseases and nutritional deficiencies based on plant coloration.” This is merely a simplified experimental substitute. If possible, the author should test on a real plant disease dataset.

We appreciate your suggestion, but we see this as a future direction for our research. The primary goal of the current work is to identify "Good Plant," "Sick Plant," and "No Plant" using computer vision in a purely illustrative manner. The amount of yellow on a green plant could indicate the degree of infection, which could then be used to determine the appropriate amount of pesticide to be applied. In other words, the proposed system could incorporate a variable application rate. However, this approach has been categorized as future work and is outside the scope of the current proposal.

 

In Figure 5, the author shows changes in load weight. Are these results derived directly from pump switch data? Are such predictions accurate? Has the author compared them with ground truth?

To clarify how we conducted the calibration process, the following comment has been added to the end of Subsection 2.1:

The load weight measurement system was calibrated using a commercial precision scale as a reference, which has a sensitivity of 1 gram and a maximum capacity of 10 kg. The calibration process involved adjusting the load cell signals relative to the reference scale, based on multiple readings for weights ranging from 0 to 1 kg, both ascending and descending. This method enhances the accuracy and reliability of the measurements obtained by the system. In summary, this procedure ensures that the values obtained from the load cell accurately reflect the weight in grams of the liquid contained in the tank.

 

In Figure 6, the author outlines the overall experimental process, including UAV flight paths and corresponding plant disease classifications. I suggest moving this figure earlier in the paper, before the experimental results section, to provide readers with a clearer understanding of the experiment's process. Additionally, the author should clarify in both the figure and the text the advantages of this method and how it differs from current research.

Thanks for your suggestion, we improve it in the revised manuscript.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The article has been greatly improved.