Gas-Driven Endoscopic Robot for Visual Inspection of Corrosion Defects Inside Gas Pipelines

Round 1

Reviewer 1 Report

 In this work, the authors present the design, validation, prototype, and preliminary experiment of a robot for gas pipe internal surface inspection. Overall the manuscript is easy to follow and the mechanical design and working principles are well explained. However, the current manuscript lacks several essential components for publication. 

1. experiment design: The robot is for gas driven pipe inspection. However, during the testing, the prototype was manually pulled through a short section of pipe. This is very limited: the two-bowl design is critical for the locomotion and yet it is not tested with real pressured gas. That is the whole point of the design: gas driven yet the prototype was never tested using pressurized gas. This can be tested by testing the prototype through a section of pipe( no need to be super long just a short demo one of 3-4 meters should suffice) and pressurizing one end of the robot to test its locomotion (e.g., compressed air). This is essential to validate the two-bowl design and validate if the speed adjustment design actually works. Without this part, the manuscript is incomplete. 

2. The data fusion algorithms should be explained in detail. The three encoders will have different readings when passing the curved section. How will the three sets of data be fused? 

3. Typo on line 197 Adam or damas?

4. Please provide and discuss method for odometer data drifing. The wheel may easily slip or spin less than it should if the surface contact has diff coef of friction (contamination and debria on the inner surface of the pipe). Over long pipeline, such error can accumulate and compropise the accuracy of the odometer and defect location.

Author Response

In this work, the authors present the design, validation, prototype, and preliminary experiment of a robot for gas pipe internal surface inspection. Overall the manuscript is easy to follow and the mechanical design and working principles are well explained. However, the current manuscript lacks several essential components for publication.

1. experiment design: The robot is for gas driven pipe inspection. However, during the testing, the prototype was manually pulled through a short section of pipe. This is very limited: the two-bowl design is critical for the locomotion and yet it is not tested with real pressured gas. That is the whole point of the design: gas driven yet the prototype was never tested using pressurized gas. This can be tested by testing the prototype through a section of pipe( no need to be super long just a short demo one of 3-4 meters should suffice) and pressurizing one end of the robot to test its locomotion (e.g., compressed air). This is essential to validate the two-bowl design and validate if the speed adjustment design actually works. Without this part, the manuscript is incomplete.

Author response: Thanks very much for your valuable comments. Indeed, as stated by the reviewer, the validity of the designed endoscopic robot must be driven by gas pressure to be finally verified. The traction tension test can be used to check the robot's passing performance before this. But, the speed adjustment function cannot be completed by traction test. Unfortunately, we did not complete the gas-driven experiments before this paper was submitted.

Fortunately, we completed a gas-driven experiment with an actual gathering pipe before the deadline. Although not perfect, the prototype robot's passing ability is still well demonstrated.  The bowl plays the role of traction and support in this robot structure. Since it is overfilled in the pipe diameter, it can be pushed by the gas with a certain pressure, and the speed regulation device actually regulates the speed by adjusting the amount of gas passing through.

Therefore, this paper verifies the passing ability of this robot by simulation, traction and gas drive tests.

According comments we have added some descriptions of the experimental results regarding the gas drive in last 2 sections, as follows.

In addition, initial testing of the pass ability of the gas-driven robot on an actual gathering pipeline was completed using the pig receiver and launcher at the gas gathering station. The length of the pipeline, i.e. the distance between the pig receiver and launcher, is approximately 5.75 km. The pressures at the inlet and outlet are 2.3 and 2.2 MPa, respectively. The total running time is about 28 minutes. Fig. 19 shows that both the traction unit and the prototype can pass through the target pipeline smoothly under gas drive. However, because the pipeline was not cleaned several times before the test, a large amount of oily dirt was still present in the pipeline, as can be seen in Fig. 19(c) for the traction unit after passing through the pipeline compared to the other two units before the test. This caused the distance measurement sensors to fail.

Furthermore, an actual gas drive test has been completed on a 5.75 km long gathering pipeline. Although the internal conditions of the experimental pipeline limited the effectiveness of the distance sensor and image sensor measurements, the passing ability was fully confirmed. Further research work includes distance measurements based on multi-sensor fusion and sensing data post-processing.

2. The data fusion algorithms should be explained in detail. The three encoders will have different readings when passing the curved section. How will the three sets of data be fused?

Author response: Thanks very much for your valuable comments. It is true that the odometer may slip during the movement and may also be suspended in the bend, which will cause it not to record the displacement accurately. An easy way to do this is to keep the longest recorded sensor results. In fact, the odometer should not be the only sensor, it also should include inertial navigation system, image, etc.. A fused positioning method based on multi-sensors fusion will be included in the final product. Related works in data processing are performed offline and are out of the scope of this paper for now.

3. Typo on line 197 Adam or damas?

Author response: Thanks very much for your valuable comments. We have revised it.

4. Please provide and discuss method for odometer data drifing. The wheel may easily slip or spin less than it should if the surface contact has diff coef of friction (contamination and debria on the inner surface of the pipe). Over long pipeline, such error can accumulate and compropise the accuracy of the odometer and defect location.

 

Author response: Thanks very much for your valuable comments. Indeed, situations such as slippage of the meter counting wheel during the motion can lead to errors. Therefore, it is not appropriate to rely on the meter wheel alone for position recording. We fuse multi-sensors fuse to improve the positioning accuracy, which will be covered in the subsequent study.

Reviewer 2 Report

1. In this paper, the authors developed a new robot called GDPER that consist of driving, odometer and visual modules in simulation environment and tested in real system. This robot can able to move in the long distance natural  gas pipeline. The robot has no electric motors as well.

    

   • Manuscript is well written and easy to read and understand.
   • Proposed work is precisely written with detail architecture.

    

   • The robot kinematic model and its effect on the movement of the robot is not clear. It would be better if the author could summarize and analyze them to show the motivation and innovation of their proposed system.
   • Some details about the developed  robot need to be included, like how to generate its trajectory in the pipeline, sensitivy of the measurement from the sensor and their analysis. It would be better if the authors could provide more details.
   • Also, there is no inference speed or computational trajectory analysis. It would be better if the authors could provide some position or velocity analysis in real environment. More analysis and comparison of the existing works can be added.

Author Response

Thanks very much for your valuable comments. Indeed, a key factor for the endoscopic robot to reach its final inspection goal is the ability of the sensors to obtain accurate measurements. These efforts are what the authors must further investigate subsequently. To validate the passing ability of the prototype, we have added a description of the experimental results of gas drive in the actual gathering pipeline. Please see the highlighted texts in the revised version.

Round 2

Reviewer 1 Report

Thanks for addressing the comments. It is great to see visual results of an actual gas-driven locomotion testing and the mechanism was shown to be effective. Figure 19 actually helps readers appreciate the complex working condition of the collection pipe inspection robot. Is it a norm/standard protocol to clean the pipe several times before the inspection? Otherwise the dirt can cover the defect easily. Please include a discussion around this. 

Author Response

Thanks very much for your valuable comments. As you have pointed out, it is a requirement of the relevant specifications to clean the pipeline several times by the pig before performing the internal inspection to achieve the internal inspection operating conditions.

Since the internal condition of the pipeline is not known before the relevant detector conducts inspection operations, it can make operating the detector subject to many uncertainties.

Dirt in the natural gas gathering pipeline can lead to a reduction in the inner diameter of the pipeline and an increase in the frictional resistance, resulting in jamming of the detector, bias grinding and other conditions. Therefore, when running the detector, it is necessary to make pigging operations in advance, which can improve the inner wall finish of the pipeline, improve the pipeline transmission capacity and extend the service life of the pipe, and the effect of pipeline clearing is directly related to the accuracy of the detector's detection data.

This is especially important for visual inspections, as dirt in the pipeline can cover or mask defects, leading to errors in judgment or misjudgments.

Since this experiment was conducted on a real natural gas gathering pipeline, multiple pigging operations may affect the amount of natural gas transmitted, so multiple pigging operation were not performed. Therefore, in order to obtain a clear image of the inner wall of the pipeline, subsequent experiments should be performed in accordance with the relevant specifications.

According comments we have added a description of pigging operation. Please see the highlighted texts in the revised version and as follows.

This suggests that, in order to assure the accuracy of the data, multiple pipeline pigging operations must also be completed in accordance with the relevant specifications for internal inspection activities before visual examination.

 

 

Reviewer 2 Report

I am satisfied with the revised version of the article, I do not request any other additions or changes.

Author Response

Thanks