Design and Testing of a Wheeled Crop-Growth-Monitoring Robot Chassis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Comments on the manuscript can be found in the attached pdf file

Comments for author File: Comments.pdf

Author Response

Dear reviewer:

Thank you very much for your thorough review and relevant insights. We believe you have accurately grasped the essence of this article. During our research, we encountered a scarcity of valuable references related to agricultural robot chassis in machinery design. As a result, we are eager to disseminate comprehensive design details of our high-clearance wheeled robot chassis. Our intention is to offer valuable insights that can potentially benefit the research efforts of others in this field.

Here are our detailed responses to your questions and suggestions.

To the General and conceptual considerations on the manuscript:

• Point 1, the title could be changed as Robot Chassis and the content related to future upgrades like electric stepper motors, guidance or self-awareness systems for obstacle avoidance or any other sort of navigation system should be added.

Our initial objective was to create a fully self-designed agricultural robot platform, encompassing the chassis, robotic arms, and sensors. Consequently, we initially named it "Robot." However, this manuscript predominantly focuses on the chassis design aspect. Therefore, it would be more fitting to refer to it as the "robot chassis." Furthermore, we have included a section in the conclusion outlining potential future upgrades for the robot.

 

 

• Point 2, details about the preliminary work of picking the right type of wheels and which variables authors kept in consideration for the purpose should be added.

The selection of wheel type primarily takes into account the row spacing used in conventional wheat cultivation methods, which typically falls within the range of 10 to 35 cm. It is crucial to ensure that the wheel's width does not exceed the planting row spacing to minimize the risk of wheat damage. As a result, we opted for wheels with a width of 7 cm.

These wheels are equipped with hub motors, which are directly integrated inside the wheel. This design choice serves the purpose of preventing wheat leaves from getting entangled with the motor during the middle and later stages of wheat growth, thus reducing the potential for motor failure.

We have incorporated this pertinent information into the manuscript.

 

• Point 3, differences in robot weight between simulation and actual use, especially the use of ballasts.

As the chassis was designed with a focus on lightweight construction and the spectral sensor testing necessitated a relatively stable environment, we did not factor in stability under high-speed movement conditions. Consequently, we did not employ any ballasts in practice.

This information has been included in the manuscript.

 

• Point 4, A section which focuses on the limits of the proposed chassis should be included to specify how different slopes, environmental conditions, terrain conditions and payloads would affect the machine.

The question raised in Line 443 aligns with the underlying rationale, and we have provided a comprehensive response below. Furthermore, we have incorporated the reasons for conducting obstacle simulation tests into the introductory paragraph of section 2.2.2.

 

This information has been integrated into the manuscript.

To specific aspects that need to be clarified or questions on the manuscript:

• Line 4: one author is missing in the list or remove “and …”;

Sorry for this mistake.

 

• Line 11: address of the corresponding author is missing;

We have added it.

 

• Line 106: Ackermann steering vehicles should also be mentioned for the same topic of the sentence. I can suggest you this interesting paper from an EU Horizon project: R.F. Carpio et al., “(Carpio et al., 2020)” (DOI: https://doi.org/10.1109/LRA.2020.2967306) published in the journal IEEE Robotics and Automation, vol.5 issue 2;

Great suggestion and it should be, we have added it.

Line 207: for what type of field conditions the tyre-soil rolling friction is estimated? What is the standard soil condition considered for the test?

We selected local loamy soil as the research object. The rolling friction coefficient range of loamy soil is about 0.3-1.9. We took 0.8 as the reference value.

 

Line 371: what is the dimension of pits and bumps that have been created on the field? That information should be included;

 

In our research, we initially aimed to assess soil data under real-world conditions. However, a significant disparity in soil compactness exists between man-made obstacles (such as pits and bumps) and obstacles encountered in natural settings. This dissimilarity makes it challenging to establish a direct correlation with simulation results. Hence, the simulation conditions we established essentially represent the most challenging obstacles that the robot might face under actual field conditions. The underlying principle is that if the robot chassis can navigate these simulation conditions seamlessly, it should theoretically be capable of performing well in real-world scenarios.

These explanations have been included in the manuscript for clarity.

• Line 433: discussion regarding the effects of wheel pressure on soil should also be mentioned;

In our subsequent research, we conducted targeted experiments to investigate the influence of soil conditions, particularly concerning wheel pressure, on the robot chassis's running speed. This endeavor unveiled a relatively intricate issue, and we plan to publish a separate paper dedicated to this specific topic. Below, we provide some of the test results and conclusions we aim to elucidate:

Our primary focus was on the measurement of three soil parameters that could significantly impact wheel speed, namely, straw coverage, soil compactness, and soil moisture content. We then proceeded to compare the predefined movement speed with the actual feedback from the motor.

Table 1 Moving speed test of the wheeled robot

soil compaction stands out as the most influential factor impacting the running speed of the robot chassis. Notably, as soil compactness increases, the power demand of the hub motor decreases when achieving the same running speed. Additionally, straw coverage and moisture content also exert their effects on speed during real-world operation.

Upon analyzing actual test conditions, we observe that when the straw coverage on the wheat field road is high, the contact area between the tires and the ground becomes smaller, while the contact area with the straw becomes larger. Given the smoother surface of straw compared to soil, it facilitates the operation of the robot chassis. Consequently, the wheeled robot chassis requires less power under these conditions.

Conversely, when soil moisture content is high and straw coverage is low, the contact area between the tires and the soil expands. The soil's moisture and stickiness create substantial resistance for the robot chassis, making it challenging to operate and necessitating greater power for propulsion.

 

• Line 481: authors could provide some data from simulations like power required for the selected payload and hypothetical battery consumption for a standard crop monitoring activity.

The specifications of the battery source we utilized were 60 V with a capacity of 20 A/h. Each hub motor operated at full load with an approximate power rating of 480 W (48 V, 10A). In our real-world testing scenario, the chassis carried a payload of approximately 15 kg, which comprised 2 kg for the laptop, 4 kg for the battery, and roughly 9 kg for the sensors and the sensor support. The test speed was configured to range from 0.1 m/s to 0.3 m/s, and the chassis exhibited a continuous operational capability for approximately 3 hours.

These details have been incorporated into the manuscript for clarity and completeness.

Reviewer 2 Report

Comments and Suggestions for Authors

The authors propose an agricultural robotic system for monitoring crop growth. The system consists of two main subsystems that are described in more detail, the robot mechanism and the sensory system for the analysis of crop growth. The control subsystem is only briefly mentioned. The system was validated in simulation and real environment.

The topic is of interest. However, the paper is not well organized and difficult to read. Language needs significant improvement, and sentences must often be more clearly written. Most of the figures are not informative and need significant improvement.

Before a thorough review of the manuscript can be performed, authors need to improve the structure of the manuscript, improve technical quality, provide more details about their approach, better discuss their validation, improve figures, and significantly improve language. Here are some examples of things that need to be improved. However, this is not a complete list; the entire manuscript must be improved.

11)      The description of the robot system is very “wordy” without references to the figures. There are too many details about bolts and holes, but relevant information about the robot mechanism is not available or difficult to comprehend. There is almost no information about traction and steering actuation (motors, gears, torque, etc.); more details about the suspension system must be provided … Leave out details about threaded connections. Provide adequate details in Figures 3 and 4. Use references in the text to link descriptions with annotations in figures.

22)      The stress analysis is incomprehensible and probably unrealistic as well. It is impossible that the deformation would be as small as 1.39xe-10 mm (this is a subatomic scale). Stress values provided in MPa are not informative. Figure 5 is useless. What could be relevant is the deformation of the robot chassis and the wheel support legs. However, I am not sure if the authors can provide realistic values. Probably, the entire section on stress analysis could be left out.

33)      The dynamic simulation of the robot system could be interesting (by the way, the title of chapter 2.2.2 should be changed to “dynamic”). However, more details need to be provided about simulation parameters (e.g. what does it mean “soft soil”). From the description, it is unclear if the simulation results are realistic (e.g. sudden changes in acceleration during obstacle crossing). Images of the robot in a simulation environment are not informative. It would be better to provide a path profile in the plot that shows robot displacements. Also, the plot needs to be improved to make the text more readable (the plot should also be in vector format).

44)      The sensing unit described in Chapter 2.3 could be interesting and relevant for readers. However, the description of the custom-developed sensing unit is very superficial. Much more detail needs to be provided (of-the-shelf measurement systems used for comparison are described in much more detail). The reader should get a complete picture of the sensory system, as this is one of the main contributions of the manuscript.

55)      The control system is also very vaguely described. However, as this is not the main contribution of the manuscript, this section does not need to be very long. Just provide more details about localization and navigation approaches.

66)      The protocol for system validation is mostly suitable. However, the description needs to be improved, and discussion about the obtained sensory results would be welcomed.

77)      This is not a complete list of manuscript shortcomings. I encourage authors to improve the entire manuscript in terms of writing and providing relevant information for the reader. 

 

Comments on the Quality of English Language

Language needs significant improvements.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Authors' reply fully satisfies the modification requests from review round 1.