Assessing Traffic-Flow Safety at Various Levels of Autonomous-Vehicle Market Penetration

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The article is well done and significant. The theme of the article has a practical value. The advantage of the article is a good presentation of interesting computation simulation with genetic algorithms. The disadvantage (in my opinion) is that the article is written for people who are fluent with transportation problems.

There are such recommendations:

1. In abstract it would be better to describe what is VISSIM, before it mentioned (see line 10).

2. In Introduction it would be better to described about criteria to divide vehicles into levels and described what does it means that autonomous vehicles belongs to level 2 or 3 (see lines 35-37)

3.  The concept IDM should be define as Intelligent Driving Model in line 56 not in 75.

4. The Wiedemann model and other models (see line 75) should have the links for articles where they have introduced.

5. It is unclear the difference about  the Wiedemann 99 and Wiedemann 74 models (see lines 78-80).

6. It would be better to add a link to the used data (see line 107).

7. It should be word "VISSIM" instead "VIS-SIM" in line 157.

8. In my opinion, it would be better  in Table 1 to add a reference number to mentioned publications, for example Parket Ohet al.(2006) [12].

9. It is unclear why parameters CC2, CC4, CC5, CC6 in Table 3 are equal to zero. Especially,  CC2 car-following distance (m) ?

10.It would be better to organize  Table 7 in such way, that eliminates the similar columns with number 111.

11. It would be better to add some plots for  illustrating the simulation.

12. It is unclear why the authors are investigated level 2-3 (see line 37) when  simulation shows AV's level 1 (see line 380)

After minor revision the article can be published.

 

 

 

Author Response

Dear Reviewer,

Thank you for your valuable feedback, which has significantly contributed to the improvement of our research.

We have carefully reviewed your comments and have prepared the following responses:

1. In abstract it would be better to describe what is VISSIM, before it mentioned (see line 10).
    : VISSIM is a representative microscopic traffic simulation used for analyzing traffic flow and safety. This explanation has been added to lines 12-13.
   2. In Introduction it would be better to described about criteria to divide vehicles into levels and described what does it means that autonomous vehicles belongs to level 2 or 3 (see lines 35-37)
    : Information on SAE International's classification criteria and the technological level of autonomous vehicles in 2023 has been added to lines 28-30.
   3. The concept IDM should be define as Intelligent Driving Model in line 56 not in 75.
    : Your feedback has been incorporated into lines 70 and 52.
   4. The Wiedemann model and other models (see line 75) should have the links for articles where they have introduced.
    : Reference 15 has been revised to include a link to the paper introducing the Wiedemann model and other models (see lines 439-440).
   5. It is unclear the difference about the Wiedemann 99 and Wiedemann 74 models (see lines 78-80).
    : The differences between the Wiedemann 99 and 74 models have been clarified by adding their application scope and functional differences in lines 73-81.
   6. It would be better to add a link to the used data (see line 107).
    : The address and source for accessing the data have been added in lines 111-115: "The data is publicly available and can be accessed at http://doi.org/10.21949/1504477."
   7. It should be word "VISSIM" instead "VIS-SIM" in line 157.
    : The wording in line 157 has been revised.
   8. In my opinion, it would be better in Table 1 to add a reference number to mentioned publications, for example Parket Ohet al.(2006) [12].
    : Reference numbers have been added.
   9. It is unclear why parameters CC2, CC4, CC5, CC6 in Table 3 are equal to zero. Especially, CC2 car-following distance (m) ?
    : A reference has been added in line 219, stating that CC2, CC4, CC5, and CC6 are variables reflecting human behavior and are calibrated to 0 when used as parameters for autonomous vehicles.
    : CC2 is a value that sets the range of how much additional following distance a driver will maintain compared to the desired safe distance before approaching the vehicle ahead. It was determined that this distance is unnecessary for autonomous vehicles.
   10. It would be better to organize Table 7 in such way, that eliminates the similar columns with number 111.
    : An error in the result value of 111 has been identified and corrected.
   11. It would be better to add some plots for illustrating the simulation.
    : The presentation of simulation results in figures has been omitted as it duplicates the content of Tables 7 and 8 and requires repetitive expression of graphs with the same meaning.
   12. It is unclear why the authors are investigated level 2-3 (see line 37) when simulation shows AV's level 1 (see line 380)
    : The statement in line 380 refers to risk level 1, while levels 2-3 in line 37 refer to the current technological level of autonomous vehicles.

We hope that our responses adequately address your concerns. We are grateful for your time and expertise.
Sincerely,

Dr. Jun Tae Park

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

1) The study titled "Assessing Traffic Flow Safety at Various Levels of Autonomous Vehicle Market Penetration" presents a thorough and insightful analysis of the impacts of increasing autonomous vehicle adoption on traffic flow safety. The authors have utilized the well-established simulation tool VISSIM and applied advanced methodologies, such as genetic algorithms, to optimize input variables, ensuring a high level of accuracy and reliability in their results.

 

2) The study's innovative approach to simulating different levels of autonomous vehicle penetration and analyzing their effects on traffic safety is commendable. The detailed examination of various scenarios, including the introduction of autonomous lanes, provides valuable insights into potential safety measures that could be implemented as the adoption rate of autonomous vehicles increases.

 

3) One of the key strengths of this research is its comprehensive data analysis and the use of real-world data from the US-101 highway. This adds significant credibility to the findings and offers a practical perspective on the real-world implications of autonomous vehicle integration.

 

4) Additionally, the study highlights important considerations for future research, such as the need to reflect actual driving behaviors of autonomous vehicles and to explore the effects on different road types. This forward-looking approach is essential for developing a more holistic understanding of autonomous vehicle impacts and improving traffic safety in diverse environments.

 

5) This study makes a significant contribution to the field of transportation engineering and autonomous vehicle research. It offers valuable guidance for policymakers and practitioners on the safe and efficient integration of autonomous vehicles into existing traffic systems. The meticulous methodology, robust analysis, and practical recommendations make this a standout study in the domain.

Comments on the Quality of English Language

English is good for publishing.

Author Response

Dear Reviewer,

Thank you for your valuable feedback, which has significantly contributed to the improvement of our research.

We have carefully reviewed your comments and have prepared the following responses:

Thank you for your positive and forward-looking feedback on our research results. We have revised our methodology and analysis sections based on your valuable comments.

We hope that our responses adequately address your concerns. We are grateful for your time and expertise.

Sincerely,

Dr. Jun Tae Park

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript focused primarily on analyzes the safety of traffic flow due to changes in the introduction rate of autonomous vehicles using the driving behavior of general vehicles analyzed based on actual data and the input variables of autonomous vehicle driving behavior used in previous studies for sections that include highway grade separations, and examines the operation of lanes dedicated to autonomous vehicles.

I found article interesting. Transport Safety and Auonomus Vehicle using is very relevant today. The methodology is well explained. The title of the paper " Assessing Traffic Flow Safety at Various Levels of Autonomous Vehicle Market Penetration"  reflects its content. Possible future research described.

The implementation needs some work: the document must be carefully checked as there are inaccuracies and style errors. Listed below are some comments:

1. Abstract:

The research design is inappropriate. Firstly, the familiar and unscientific style needs to be changed.

Eg. "In this study, we optimized .." must to be changed “This article optimizes /discusses.

2. Abstract in general look like conclusions. I recommend writing here the objectives of the study, not the results.

3. Lines 44-48 and 49-53 are same.

5. Line 57. Abriviation IDM only explained in line 75 Intelligent Driving Model (IDM).

6. Line 157. VIS-SIM

7. Line 186. Abriviation GEH not explained

8. Line 235. Assess-ment

Author Response

Dear Reviewer,

Thank you for your valuable feedback, which has significantly contributed to the improvement of our research.

We have carefully reviewed your comments and have prepared the following responses:

1. The abstract is too long and verbose
 : The abstract has been revised in lines 10-24 to remove unnecessary details and concisely convey the main topic.
2. There are several typos
 : Typographical errors throughout the text have been corrected.
3. Table 4 has Km/s as unit of measure rather than Km/h
 : Units have been revised.
4. Tables should be formatted with the same font of the text and should be clear, self-explaining. Too much confusion within a table
 : Complex table contents have been organized, and fonts have been unified.
5. The introduction of the SSAM analysis should be addressed in the methodology
 : Section 3.6, "Selection of Evaluation Metrics," has been added to lines 221-243 to explain the SSAM analysis.
6. The methodology section misses important pieces of info about the steps used and about the reason to use randomized parameters
 : Various references were cited to select variables used in traffic flow simulations to simulate real vehicle behavior in the program, as shown in lines 158-160. The results were applied to the genetic algorithm as minimum and maximum values.
7. A sensitivity analysis lacks
 : The purpose of this study is to propose a methodology for analyzing the safety of traffic flow according to the introduction rate of autonomous vehicles (Lv. 4-5). The analysis scope includes the number of conflicts and average travel speed, which are indicators of traffic flow safety for each autonomous vehicle introduction rate. Additional sensitivity analyses, applying various scenarios and geometric conditions, will be considered in future research, as reflected in the conclusion (lines 398-406).
8. There are few details about the road selected and the reason to choose that specific road, that specific time interval, and so fort. More details are required to justify the simulation basis
 : Information on vehicle information provided every 0.1 seconds for all vehicles on the road and the suitability of using US I-101 highway data due to its auxiliary lanes allowing free lane changes has been added to lines 106-115.
9. Important details and considerations about the autonomous lanes and their implications on road geometry, road traffic and new driving scenarios miss. But it is still a good idea for analysis.
 : Regarding the impact analysis of road geometry, it is expected that the introduction rate of autonomous vehicles will continue to increase. Changing the road geometry every time the introduction rate increases would be very costly. Therefore, a simple way to separate regular and autonomous vehicles is to introduce dedicated lanes for autonomous vehicles. These lanes can be operated similarly to existing bus lanes, making them easy to implement. Since the road geometry would not change with the introduction of dedicated lanes, no additional measures are needed.
 : Regarding the overall impact on road traffic, lines 345-349 have been added to state that operating dedicated lanes for autonomous vehicles is suitable for introduction rates of 10-30%, based on safety and operational efficiency.
10. Avoid cluster of references, e.g. 1-5 but explain the main outcome you are citing
 : References have been used individually to explain each main result and are presented in lines 30-33.
11. Please improve the literature used as highlighted before.
 : Based on your feedback, references have been supplemented and reorganized.

We hope that our responses adequately address your concerns. We are grateful for your time and expertise.

Sincerely,

Dr. Jun Tae Park

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The study has some novelties if compared to the existing literature, as the choice of investigating the autonomous lane. All the rest of the manuscript overlaps recent studies on this topic, which are not cited in the introduction. Moreover, literature reviews about the most suitable simulator to use for safety assessments with AVs exist  and are not cited too.

In the study there is a lack of context about the choice of the simulator, the boundary condition of the experiment and aspect of novelties of the proposed research. These aspects represent major concerns to address.

Below some minor suggestions:

-The abstract is too long and verbose

-there are several typos

- Table 4 has Km/s as unit of measure rather than Km/h

- Tables should be formatted with the same font of the text and should be clear, self-explaining. Too much confusion within a table

- The introduction of the SSAM analysis should be addressed in the methodology

- The methodology section misses important pieces of info about the steps used and about the reason to use randomized parameters

- A sensitivity analysis lacks

- There are few details about the road selected and the reason to choose that specific road, that specific time interval, and so fort. More details are required to justify the simulation basis

- Important details and considerations about the autonomous lanes and their implications on road geometry, road traffic and new driving scenarios miss. But it is still a good idea for analysis.

- Avoid cluster of references, e.g. 1-5 but explain the main outcome you are citing

- Please improve the literature used as highlighted before.

Comments on the Quality of English Language

Good enough, but some parts need to be modified.

Author Response

Dear Reviewer,

Thank you for your valuable feedback, which has significantly contributed to the improvement of our research.

We have carefully reviewed your comments and have prepared the following responses:

1. The abstract is too long and verbose
 : The abstract has been revised in lines 10-24 to remove unnecessary details and concisely convey the main topic.
2. There are several typos
 : Typographical errors throughout the text have been corrected.
3. Table 4 has Km/s as unit of measure rather than Km/h
 : Units have been revised.
4. Tables should be formatted with the same font of the text and should be clear, self-explaining. Too much confusion within a table
 : Complex table contents have been organized, and fonts have been unified.
5. The introduction of the SSAM analysis should be addressed in the methodology
 : Section 3.6, "Selection of Evaluation Metrics," has been added to lines 221-243 to explain the SSAM analysis.
6. The methodology section misses important pieces of info about the steps used and about the reason to use randomized parameters
 : Various references were cited to select variables used in traffic flow simulations to simulate real vehicle behavior in the program, as shown in lines 158-160. The results were applied to the genetic algorithm as minimum and maximum values.
7. A sensitivity analysis lacks
 : The purpose of this study is to propose a methodology for analyzing the safety of traffic flow according to the introduction rate of autonomous vehicles (Lv. 4-5). The analysis scope includes the number of conflicts and average travel speed, which are indicators of traffic flow safety for each autonomous vehicle introduction rate. Additional sensitivity analyses, applying various scenarios and geometric conditions, will be considered in future research, as reflected in the conclusion (lines 398-406).
8. There are few details about the road selected and the reason to choose that specific road, that specific time interval, and so fort. More details are required to justify the simulation basis
 : Information on vehicle information provided every 0.1 seconds for all vehicles on the road and the suitability of using US I-101 highway data due to its auxiliary lanes allowing free lane changes has been added to lines 106-115.
9. Important details and considerations about the autonomous lanes and their implications on road geometry, road traffic and new driving scenarios miss. But it is still a good idea for analysis.
 : Regarding the impact analysis of road geometry, it is expected that the introduction rate of autonomous vehicles will continue to increase. Changing the road geometry every time the introduction rate increases would be very costly. Therefore, a simple way to separate regular and autonomous vehicles is to introduce dedicated lanes for autonomous vehicles. These lanes can be operated similarly to existing bus lanes, making them easy to implement. Since the road geometry would not change with the introduction of dedicated lanes, no additional measures are needed.
 : Regarding the overall impact on road traffic, lines 345-349 have been added to state that operating dedicated lanes for autonomous vehicles is suitable for introduction rates of 10-30%, based on safety and operational efficiency.
10. Avoid cluster of references, e.g. 1-5 but explain the main outcome you are citing
 : References have been used individually to explain each main result and are presented in lines 30-33.
11. Please improve the literature used as highlighted before.
 : Based on your feedback, references have been supplemented and reorganized.

We hope that our responses adequately address your concerns. We are grateful for your time and expertise.

Sincerely,

Dr. Jun Tae Park

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Most of the errors have been corrected. But the authors response are not according to my review. The document must be carefully checked as there are inaccuracies and style errors.

Lines 38....42 and 43.....47 are same.

Abriviation GEH not explained

Line 233. Assess-ment

Author Response

Dear reviewer

Thank you for giving me a good opinion to make it a high-quality paper.

We have carefully reviewed your comments and have prepared the following responses:

1. The document must be carefully checked as there are inaccuracies and style errors.
 : Thank you for your thorough review of my paper. Based on your feedback, I have reviewed the overall content and revised the writing style.

2. Lines 38....42 and 43.....47 are same.
 : I have also removed any redundant sentences.

3. Abbreviation GEH not explained
 : GEH is not an abbreviation, but a person's name. The GEH formula gets its name from Geoffrey E. Havers.(Line 180)

4. Line 233. Assess-ment
 : I have corrected a typo.(Line 229)

We hope that our responses adequately address your comments. We are grateful for your time and expertise.
Sincerely,

Dr. Jun Tae Park

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The suggestions have been addressed. I just suggest to still improve the literature used because studies on this topic can help you to highlight the relevance of your study.

Author Response

Dear reviewer

Thank you for giving me a good opinion to make it a high-quality paper.

We have carefully reviewed your comments and have prepared the following responses:

1. I just suggest to still improve the literature used because studies on this topic can help you to highlight the relevance of your study.
 : We have added some references that can complement the references we have used to achieve improved research results, and we have also added them to the references. (Lines 58-60, 30, 429-432, 435-436, 450-451)

We hope that our responses adequately address your coments. We are grateful for your time and expertise.
Sincerely,

Dr. Jun Tae Park

Author Response File: Author Response.pdf