Improvement Practices in the Performance of a CPS Multiple-Joint Robotics Simulator

Round 1

Reviewer 1 Report

The authors in the article presented the issue related to CPS Multiple-Joint Robotics Simulator. To solve the problem, a visual approach was proposed to increase the complexity of the code based on static analysis that identifies bad code in terms of performance. Extracting the complexity of CPS software by visualizing the code and identifying high complexity involves measuring performance for repair. The benefits of the presented work were not clearly shown in terms of comparison, effectiveness and results achieved. The article requires a slight stylistic correction, no reference in the text to Fig. 3. Figures 1 and 4 are difficult to read.

Author Response

The authors in the article presented the issue related to CPS Multiple-Joint Robotics Simulator. To solve the problem, a visual approach was proposed to increase the complexity of the code based on static analysis that identifies bad code in terms of performance. Extracting the complexity of CPS software by visualizing the code and identifying high complexity involves measuring performance for repair.

1) The benefits of the presented work were not clearly shown in terms of comparison, effectiveness and results achieved.

- In Table 2, table 3,  we mentioned the improved complexity values and performance of the code modules between the before refactoring and the after refactoring. (in 9~10 pages)

2) The article requires a slight stylistic correction, no reference in the text to Fig. 3.

- we did fix them.(in 5 page)

3) Figures 1 and 4 are difficult to read.

- we did fix to read clearly figures.(Figure 1 -> 3 page, Figure 4 -> 6 page)

 

Author Response File: Author Response.docx

Reviewer 2 Report

I understand and agree the significance of the paper on improvement of the software performance within the field of smart devices or systems. Still, I don't get why the proposed methods could not be applied on other kinds of software besides the CPS.

Please, provide more details and particularities (if any) of the proposed software performance measurements on CPS or other SMART systems in order to emphasise them better. Otherwise, I think the performance results (e.g. time) presented by the paper could be widely scaled among the software systems having similar architectures.

Author Response

I understand and agree the significance of the paper on improvement of the software performance within the field of smart devices or systems. Still, I don't get why the proposed methods could not be applied on other kinds of software besides the CPS.

- In current Korea, many researchers are focusing on the autonomous robots simulators and vehicles in 4^th industrial revolution period. They are just developing their hardware robots very well without software quality. Therefore we are considering how to enhance the quality of software (simulators). We are also applying for low-power optimal operation on Ultra small Iot Devices to investigate and analyze the heavy method to reduce power consumption with ETRI(Electronics and Telecommunications Research Institute) in Korea.  

Please, provide more details and particularities (if any) of the proposed software performance measurements on CPS or other SMART systems in order to emphasise them better. Otherwise, I think the performance results (e.g. time) presented by the paper could be widely scaled among the software systems having similar architectures

-Now it  is very difficult to analyze huge side of software to identify some problems such as performance, power consumption, complexity, too much coupling manually. Our idea is to visualize some problems of code to easily fix them.

Author Response File: Author Response.docx

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

This paper proposed to improve code complexity of the CPS based multiple-joint robotics simulator. According to the current manuscript, the proposed approach is not specified and optimized for CPS. Based on the following review comments, this paper is suggested to be rejected and not published in this journal.

 

1. Authors showed Equation (1) to explain the software performance. However, authors measured the response time. Authors must explain why not utilize Equation (1). 

2. Since authors select the CPS simulator, authors must explain the motivation and the representative of this simulator.

3. No difference between "before" and "after" in Figure 2. Authors should explain more details of Figure 2.

4. Authors described that it displays in red if its value has above a certain score. However, authors did not explain how to define or find out the score.

5. The quality of some figures, e.g., Figures 4 and 7, is not good enough. 

6. Authors only compare the performance before and after the proposed approach. Authors are suggested to compare with other works in order to highlight the critical contribution of the proposed approach.

7. The corresponding unit should be added into experimental results, including in the tables, e.g., Tables 2 and 3, and figures, e.g., Figure 8.

8. Only one citation were published in 5 years. Authors are suggested to cite more recent papers, especially published in this journal.

9. There are some typos in the current manuscript. Authors are suggested to polish the English writing and then increase the readability of this paper.

Author Response

Authors showed Equation (1) to explain the software performance. However, authors measured the response time. Authors must explain why not utilize Equation (1).

Answer: We are software engineers who focus on code visualization for quality. Instead of availability and hardware weight, we are focusing on code size with complexity to enhance response time. Suppose if enhance complexity with code size for code quality, then we make better performance, and also better response time.

Since authors select the CPS simulator, authors must explain the motivation and the representative of this simulator.

Answer: In current Korea, many researchers are focusing on the autonomous robots (simulators) and vehicles in 4^th industrial revolution period. They are just developing their hardware robots very well without software quality. Therefore, we are considering how to enhance their software (simulators).

No difference between "before" and "after" in Figure 2. Authors should explain more details of Figure 2.

Answer: In Figure 2, we just mention how to compare between before and after refactoring. So we explain details with our robot’s simulator between before and after refactoring in Figure 7.

Authors described that it displays in red if its value has above a certain score. However, authors did not explain how to define or find out the score.

Answer: Our research is focusing on reverse engineering, which first statically analyzed with source code. We make your queries with quality factors(such as coupling), and also store the analyzed results into DB tables. For example, Coupling is composed of data, stamp, control, external, common and content coupling. Here, data coupling score is the lowest, and the content coupling score is the highest. In software engineering, the code complexity is worse than the code size, that is, the internal complex structure of software. The complexity of code size is measured by its coupling such as the calling and called modules. In this paper, we measure the coupling to assign adequate scores for each coupling. Table 1 shows the scores for each coupling.

표1. adequate scores for each coupling

Coupling	Data	Stamp	Control	External	Common	Content
Score	1	2.4	3.9	5.6	7.5	10.2

The coupling formula is as follows:

For example, the ② of Figure 5 shows the couplings between CSimulationView and CServer module, and also [CT*21] = 10.2*21= 214.0 on the directed red arrow where the number and score of content couplings(CT) are 21 and 10.2.

So, to enhance the code complexity, we define the baseline of sum of each coupling, that is, the score 30. On static analysis with source code, we automatically display the red colors of the arrow and module. Over the score 30, we do refactoring again and again until below the score 30.

The quality of some figures, e.g., Figures 4 and 7, is not good enough.

Answer: We fix them.

Authors only compare the performance before and after the proposed approach. Authors are suggested to compare with other works in order to highlight the critical contribution of the proposed approach.

Answer: Most of researchers focus on static and dynamic analysis, but Nobody focuses on code visualization with static and dynamic analysis.

The corresponding unit should be added into experimental results, including in

the tables, e.g., Tables 2 and 3, and figures, e.g., Figure 8.

Answer: We follow reviewer 1 to add all experimental results in Table 2 and 3.

Only one citation was published in 5 years. Authors are suggested to cite more recent papers, especially published in this journal.

Answer: We add more recent papers into reference.

There are some typos in the current manuscript. Authors are suggested to polish the English writing and then increase the readability of this paper.

Answer: We fix our paper.

 

 

Reviewer 2 Report

Weaknesses
========
- Abstract incomplete.
- Few related works.
- Old references.
- The work looks more like a technical report than a research paper.

Detailed Comments
=============
- In the abstract, it is not clear: the research method, the results and the conclusions.
- The paper has few related works and they are old. The most recent related work was published 7 years ago. This makes it difficult to determine what the paper's contribution is.
- In general, the references of the paper are old. 75% of the works are 8 years old or more since their publication. The information in some of the references is incomplete. For example, reference [7] is missing the year of publication.
- Figures 1 and 4 are difficult to understand. For example, in Figure 1: (i) it is not clear what the area delimited by a red line means, (ii) the meaning of the dotted line of light blue color is not clear.

Other comments
=============
- In the 1st paragraph of the introduction, the end point is missing.
- Change the word chapter to section.

Comments for author File: Comments.pdf

Author Response

1. In the abstract, it is not clear: the research method, the results and the conclusions.

Answer: In near 4th Industrial revolution environment, it will come out plenty of automatic and smart software in diverse fields such as VR/AR, Autonomous Robots and Vehicles, Smart Factoring and so on. Specially, in Korea, one of its important issues is the cyber-physical system (CPS) for monitoring and controlling the smart and automatic system in smart city. Therefore, this kind of systems need to have good performance otherwise it is too much late to work on time. To solve this, we propose a code visual approach to enhance code complexity based on static analysis, which identify bad codes against performance. To make better performance, we make our queries to identify performance degradation factors, store the statically analyzed data into DB tables, and visualize the bad operation patterns. For performance improvement, we can make refactoring with them. As a result, we enhance code complexity of the CPS-based software to get good performance. With this approach, we expect to have better performance and complexity improvement of CPS software without even power consumption.

2. The paper has few related works and they are old. The most recent related work was published 7 years ago. This makes it difficult to determine what the paper's contribution is.

Answer: We add the recent related papers into our paper. Actually we had developed this multiple robot simulator for one Robot Company at 7 year ago. At that time, we never consider of code complexity and performance. In this time, we apply our visual approach to our developed CPS simulator to enhance performance.

3. In general, the references of the paper are old. 75% of the works are 8 years old or more since their publication. The information in some of the references is incomplete. For example, reference [7] is missing the year of publication.

Answer:  We fix them.

4. Figures 1 and 4 are difficult to understand. For example, in Figure 1: (i) it is not clear what the area delimited by a red line means, (ii) the meaning of the dotted line of light blue color is not clear.

Answer: We clearly display them in Figure 1 and figure 4.

5. In the 1st paragraph of the introduction, the endpoint is missing.

Answer: We fix our paper.

6. Change the word chapter to section.

Answer: We fix our paper.

Reviewer 3 Report

The authors of the article present the operation of a multiple-pond robot simulator using a cyber-physical system. It was based on optimizing programming code to improve performance based on our code quality indicators. To determine performance improvements, it would be good to refer to the algorithms themselves and their performance. Algorithm optimization is a key element in the performance process.

Author Response

1. The authors of the article present the operation of a multiple-pond robot simulator using a cyber-physical system. It was based on optimizing programming code to improve performance based on our code quality indicators. To determine performance improvements, it would be good to refer to the algorithms themselves and their performance. Algorithm optimization is a key element in the performance process.

Answer: Thank for your opinion.

Round 2

Reviewer 1 Report

This paper proposed one code visual approach to enhance the code complexity. However, authors did not provide good enough readability and English writing. This paper is suggested to be rejected based on the following review comments.

 

1.) In the previous review, Figure 2 is suggested to be revised or explained in more details. However, authors still use the same figure without any improvement. Is it really necessary or required? 

 

2.) Figure 4 and Figure 5 are still unclear with the low resolution. However, these two figures are the most critical figures to explain the advantage or outstanding difference before and after the proposed visual approach in this paper. Authors did not provide good and detailed descriptions.

 

3.) Figure 8 in this manuscript is different from the one in the previous version. Authors should check whether this figure is correct or not. 

 

4.) Basically, this paper is not related to CPS. Authors just take the CPS software as the example to measure the performance. Authors are suggested to show the true power of code visualization. For example, how to convert the unclear modules in Figure 4 into the clear ones in Figure 5?

 

5.) English writing must be polished by native speakers. 

 

6.) Even though authors add recent works, there are still more than half citations published before 5 years.