Improving Real-Time Performance of Micro-ROS with Priority-Driven Chain-Aware Scheduling

Round 1

Reviewer 1 Report

In this paper, authors well described and developed a priority-driven chain-aware scheduling system (PoDS) based on the existing micro-ROS architecture. Although this paper is well written and easily read, while several sections in the manuscript are less informative that require more details. Here are some comments that the authors have to address.

1. Line 241: it’s suggested to add some physical pictures or circuit diagrams to help readers easy reading.

2. Lines 465 to 468: why do you choose these parameters 100 to 500 bytes to measure? Do the data packet sizes have the special effects for the performances?

3. Lines 551 to 552: llack analytical methods, llack be lack?

 Minor editing of English language required.

Author Response

Thank you very much for taking your time to review this manuscript. We have carefully read your comments and find your suggestions very valuable. Below, I will provide responses to each of your comments one by one:

Comments 1: Line 241: it’s suggested to add some physical pictures or circuit diagrams to help readers easy reading.

Responses 1: As the PoDS proposed in this paper is a software system, we were a bit confused by your suggestion to add physical pictures or circuit diagrams. However, in order to assist readers in easier reading, we have marked the detailed introduction chapters corresponding to the key components in the PoDS system in the Section 4.2 , which may be the original intention of your comment.

Comments 2: Lines 465 to 468: why do you choose these parameters 100 to 500 bytes to measure? Do the data packet sizes have the special effects for the performances?

Responses 2: We consider this comment very valuable. Firstly, the impact of the data packet size on system performance that you mentioned. Since the data transmission time is closely related to the size of the transmitted data packets, we actually intend to measure the impact of different data transmission times on the real-time performance of the system by varying the size of the transmitted data packets.

Furthermore, you mentioned why the parameter range of 100 to 500 bytes was chosen for measurement. We can roughly determine the byte range of the transmitted data packet by examining the generic robot-specific message types widely used by ROS applications. For example, the NavSatFix Message exceeds 116 bytes, and the PoseWithCovariance Message greatly exceeds 288 bytes. You can see the above at line 473 of the new manuscript.

Comments 3: Lines 551 to 552: llack analytical methods, llack be lack?

Responses 3: Yes, you are right. We have checked the entire manuscript many times and found some spelling errors and imprecise expressions. After multiple checks and revisions, we can confirm that there are no spelling or grammar errors in the manuscript. For the sake of your reading experience, we have not marked minor spelling or grammar corrections in the manuscript.

Thank you again for your time to review this manuscript. 

Reviewer 2 Report

The work is interesting, the improvement in the end-to-end latency is notable and I think deserve to published.

My main worry if it suits to Electronics or better to Robotics or  Computers (I would say the last fits better, but I left the decision to the editors)

The English requires a revision.

Author Response

Thank you very much for taking your time to review this manuscript. 

Your positive feedback on our paper's work is greatly appreciated. We believe that our work is relevant to the topic being submitted, so we are still preparing to submit this special topic. Thank you very much for your advice regarding the submission.

We have checked the entire manuscript many times and found some spelling errors and imprecise expressions. After multiple checks and revisions, we can confirm that there are no spelling or grammar errors in the manuscript. 

Thank you again for your time to review this manuscript. 

Reviewer 3 Report

The paper provides a detailed explanation of the problems inherent in the existing micro-ROS system and proposes a novel approach with PoDS. However, the following concerns need to be addressed before it can be considered for publication.

Could the authors list and explain the specific assumptions made within the modeling of callback scheduling and data transmission? This detail will help readers understand the foundational premises of the proposed system. How do these assumptions hold in varied real-world applications? It is essential to discuss whether these assumptions apply universally across different robotic systems and operating conditions. If there are scenarios where the assumptions might not hold, these should be explicitly stated.

While the conclusion summarizes the current achievements well, it would be beneficial to discuss potential future improvements or extensions to the PoDS system. Speculating on future research directions or potential new features could engage the reader and suggest how the work could evolve.

There are minor grammatical errors and some awkward phrasings throughout the text that could be revised for better readability. For instance, Line 49 ‘drivern’, Line 545  ‘probabilites’, Line 551 ‘llack’. The manuscript should consistently use either uppercase or lowercase for all abbreviations, for example, rcl, rclc in Line 76/77.

Minor editing of English language required.

Author Response

Thank you very much for taking your time to review this manuscript. We have carefully read your comments and find your suggestions very valuable. Below, I will provide responses to each of your comments one by one:

Comments 1: Could the authors list and explain the specific assumptions made within the modeling of callback scheduling and data transmission? This detail will help readers understand the foundational premises of the proposed system. How do these assumptions hold in varied real-world applications? It is essential to discuss whether these assumptions apply universally across different robotic systems and operating conditions. If there are scenarios where the assumptions might not hold, these should be explicitly stated.

Responses 1: After careful consideration of this omment , we believe the specific assumption you are concerned about is the use of only the default trigger condition ANY option mentioned in line 121. We consider your comment very valuable and thus provided a detailed explanation in line 259 for this specific assumption.

In order to better illustrate the operational behavior of PoDS, this paper selects the default trigger condition ANY option. However, this does not imply that the PoDS system proposed in this paper is restricted solely to a single trigger condition. In fact, it is compatible with all trigger condition options. In practical application scenarios, PoDS can remain consistent with micro-ROS.

Comments 2: While the conclusion summarizes the current achievements well, it would be beneficial to discuss potential future improvements or extensions to the PoDS system. Speculating on future research directions or potential new features could engage the reader and suggest how the work could evolve.

Responses 2: We think your suggestion is very correct, so we added the future work at the end of this paper. In the future, we hope to break the limitations of ROS 2 Agent by introducing the concept of embeddedRTPS design within the PoDS framework.

Comments 3: There are minor grammatical errors and some awkward phrasings throughout the text that could be revised for better readability. For instance, Line 49 ‘drivern’, Line 545  ‘probabilites’, Line 551 ‘llack’. The manuscript should consistently use either uppercase or lowercase for all abbreviations, for example, rcl, rclc in Line 76/77.

Responses 3: We have checked the entire manuscript many times and found some spelling errors and imprecise expressions. After multiple checks and revisions, we can confirm that there are no spelling or grammar errors in the manuscript. For the sake of your reading experience, we have not marked minor spelling or grammar corrections in the manuscript.

Thank you again for your time to review this manuscript.