Expressway Vehicle Trajectory Prediction Considering Historical Path Dependencies

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I appreciate the authors' focus on predicting vehicle trajectories on highways using Electronic Toll Collection (ETC) data, which is highly relevant to addressing modern transportation challenges. Highways play a vital role in transportation networks, and optimizing their efficiency is crucial for reducing congestion, enhancing safety, and fostering sustainable urban development. However, I would like to offer some suggestions for improvement:

 

1   1. The methodology and key contributions are well outlined; however, the study's objectives should be clearly articulated in the introduction for better clarity.

     2. It would be beneficial for the authors to provide insights into their data cleaning process and how anomalies in the dataset were addressed.

      3. Further elaboration is needed on how the authors constructed sequences of access points to depict the chronological order of vehicle movements.

     4. Consider moving the study's objective, currently located in section 3.1.2, to the introduction to align more cohesively with the paper's structure.

      5. Authors should explicitly describe their approach to handling missing and abnormal values during data analysis.

    6. It would be helpful for readers if the authors explained the validation process in detail.

       7. How do the results of this study align or differ from previous studies?

      8. Strengthening the link between the conclusion and the study's objective would enhance the overall coherence of the paper.

     9. Discussing the limitations of the proposed model in predicting highway vehicle trajectories, particularly for vehicles without ETC systems, would provide a more comprehensive perspective on its practical applicability.

     10. Exploring how the absence of data from vehicles without ETC systems could impact prediction accuracy, model generalization, and result representativeness would enrich the discussion on the model's performance and its real-world implications.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This paper proposes a new model to predict expressway vehicle trajectories by analyzing historical traffic patterns and leveraging ETC data. The proposed method and model are beneficial to refining intelligent transportation systems, thus enhancing traffic management through a deeper understanding of driver behaviors. The proposed method is very interesting. For instance, the historical trajectories are transformed into strings neatly and predictions can be performed thereby. The following suggestions are for reference.

1.         Please merge the “Related work(s)” into the introduction part. The introduction can be divided into “Background” and “Related works” thereafter. The subtitles in the related works looks strange, since “based on statistical/machine learning methods” is incomplete. The authors should either make “Related works” more informative, for instance, “Related works of vehicle trajectory prediction”.

2.         A nomenclature is suggested to explain all the abbreviations and symbols.

3.        Line 519, what does “a total of *** data points” stand for? Is the scale of the data points confidential?

4.         Details of training the network need to be enriched. They can be illustrated in figures or in supplementary materials.

5.         The idea of using average pooling is useful to tackle with historical trajectories with different sizes. However, other parameters may also exist that influence the vehicle trajectory, e.g., weather and road conditions. Please discuss.

6.         The nice results are based on the specific dataset. To follow the previous comment, is the generalization ability (extended to other conditions) of the trained guaranteed?

Comments on the Quality of English Language

Minor editing of English language required.  Line 519, what does “a total of *** data points” stand for? Is the scale of the data points confidential?

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The paper introduces a novel and state-of-the-art approach for pedicting vehicel trajectories on an expressway network. Specifically, in a combination with other solutions it utilizes a key concept which is also one of the main characteristics of large language models, the Word2Vec (word to vector) method in a very original way to extract the vehicle trajectories from ETC transactions registered at the different gantrys on the expressway.

As a result of the previous, the results of the paper are highly relevant and the introduced methodology can be of high interest for the other researchers in the field. The structure of the paper is also clear, its style is elegant and it fulfills the formatting requirements.

The literature review is adequately implemented. The English language of the paper is also of excellent quality.

Based on the previous, it can be stated that the paper is of very high quality from every aspect. Yet, there is still one issue that requires a minor revision, which is the fact that the question of sustainability does not appear in any form in the publication. At minimum, some remarks should be made about the possible conclusions that can be deduced from the results related to the environmental impact of the traffic flows on the expressway network. Moreover, it would be useful if the authors would describe how the proposed approach can be applied in the reduction of the environmental impact of road traffic.

After the implementation of the previous recommendation, I would recommend the publication of the paper without any other changes. 

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx