Next Article in Journal
Levels of Sustainability Awareness in Spanish University Students of Nautical Activities as Future Managers of Sports and Active Tourism Programmes
Next Article in Special Issue
Mobility-as-a-Service (MaaS) and High-Speed Rail Operators: Do Not Let the Train Pass!
Previous Article in Journal
Mission Efficiency Analysis of For-Profit Microfinance Institutions with Categorical Output Variables
 
 
Article
Peer-Review Record

A Knowledge-Based AI Framework for Mobility as a Service

Sustainability 2023, 15(3), 2717; https://doi.org/10.3390/su15032717
by Enayat Rajabi 1,2, SÅ‚awomir Nowaczyk 1, Sepideh Pashami 1, Magnus Bergquist 1, Geethu Susan Ebby 2,* and Summrina Wajid 1
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3:
Sustainability 2023, 15(3), 2717; https://doi.org/10.3390/su15032717
Submission received: 25 December 2022 / Revised: 30 January 2023 / Accepted: 31 January 2023 / Published: 2 February 2023
(This article belongs to the Special Issue Mobility as a Service: Perspectives for Large-scale Operations)

Round 1

Reviewer 1 Report

The manuscript presents the architecture and prototype of a knowledge-based system, integrating data from various sources and modules that incorporate AI techniques to make recommendations to travellers in the context of MaaS. It also implements a limited proof of concept that exemplifies its use.

It addresses the problem of mobility of interest, topicality and complexity and proposes a framework to offer solutions. The initial results and conclusions open essential research and development spaces.

Some simple issues to improve by the authors:

The manuscript presents significant advances, with a proof of concept, although it does not describe the stage of development of the complete application.

It is not clear to me to distinguish between the last three sections Results, Discussion, and Conclusions. Why not incorporate the discussion contained in the introduction and conclusion? Perhaps it is enough to introduce the claims in the results and discussion and expand the conclusions.

In the study of art, in addition to MaaS and IA, some specific research in recommendation systems, optimization, simulation and data analytics can complement it in addition to the proposed architecture and conclusions.

It would be good to describe some parameters (variables or constants) in all algorithms (some are) to better understand them. They can be added as comments in the pseudocode itself.

Some types:

Planes/routes online 47

The content of sections 5 and 7 is missing in "The structure of this paper", lines 63-66

In line 103, Capital letters after the point and follow in "in another"

Author Response

Thank you very much for the reviewer's constructive comments. Please find our responses below: 

Comment

 

The manuscript presents significant advances, with a proof of concept, although it does not describe the stage of development of the complete application.

Response

Thank you for the comment, the further focus of the project is added to the conclusion paragraph (line 380-390)

Comment

 

 

 

It is not clear to me to distinguish between the last three sections Results, Discussion, and Conclusions. Why not incorporate the discussion contained in the introduction and conclusion? Perhaps it is enough to introduce the claims in the results and discussion and expand the conclusions.

Response

Thank you for the suggestion. We merged the Results and Discussion sections as one section (Section 5). We moved a few sentences to the Introduction section and some of our findings into conclusion.

Comment

 

 

In the study of art, in addition to MaaS and IA, some specific research in recommendation systems, optimization, simulation and data analytics can complement it in addition to the proposed architecture and conclusions.

Response

Thank you for the comment. Related studies giving a different perspective of MaaS systems are added in lines 120-129

Comment

 

 

It would be good to describe some parameters (variables or constants) in all algorithms (some are) to better understand them. They can be added as comments in the pseudocode itself.

Response

Thank you for the comment. Comments and explanations were added in Algorithms 1,2,3.

Comment

Some typos:

Planes/routes online 47

Response

Thank you for the comment. Corrected (line 50)

Comment

 

The content of sections 5 and 7 is missing in "The structure of this paper", lines 63-66

Response

Thank you for the comment. Sections are added (lines 66-70)

Comment

In line 103, Capital letters after the point and follow in "in another"

Response

Thank you for the comment. Corrected (line 107)

Reviewer 2 Report

My comments are the following:
Row 65 – Before “Section 6 will discuss the challenges…”, please insert a brief description of Section 5.
Row 103 – Please insert comma after aggressive in “(in aggressive normal, and calm)”.
Row 142 – “It. The semantic” – please solve the “it” error.

Author Response

We appreciate the reviewer's positive comments. Please find our responses:

Comment: Row 65 – Before “Section 6 will discuss the challenges…”, please insert a brief description of Section 5
Response: Thank you for the comment. Sections were added (lines 66-69) ---------------------- Comment: Row 103 – Please insert a comma after aggressive in “(in aggressive normal, and calm)”.
Response: Thank you for the comment. Corrected (line 107) -------------------------------------- Comment: Row 142 – “It. The semantic” – please solve the “it” error.
Response: Thank you for the comment. Corrected (now line 156)

Reviewer 3 Report

The explanation of the framework is clear and straightforward. However, my main concern and comments are as below:

1. For the scenario presented in the implementation section, in order to provide a clearer explanation of the system, it is suggested to include screenshots of the MaaS dashboard in Section 4.

2. Algorithm 3 which was developed to recommend vehicles to travelers is just based on distance. How about other factors such as travel distance, travel time and etc.?

3. How to avoid biases in the system?

4. For the scenario presented in the implementation section, the processing part in the system is quite simple, it would be helpful if the authors can highlight the usefulness of the AI system. Additionally, it would be better if the authors can present some results in the manuscript. 

Author Response

Thanks to the reviewer's constructive comments. Please find our responses below: 

Comment

 

For the scenario presented in the implementation section, in order to provide a clearer explanation of the system, it is suggested to include screenshots of the MaaS dashboard in Section 4.

Response

Thank you for the comment. We included a screenshot of the implemented dashboard in the paper (Figure 4). For more clarity and to improve readability, we also updated Figure 5 with better labels and explanations.

Comment

 

 

Algorithm 3 which was developed to recommend vehicles to travelers is just based on distance. How about other factors such as travel distance, travel time and etc.

Response

Thank you for the comment. For the presented case study, we have considered the closest vehicles to the passenger considering their preferences. In addition, we can also consider other parameters such as the shortest travel path, the fastest mode of transport or accommodate route preferences. Updated lines 291-294. We also updated the conclusion and future works section with more insights regarding this comments.

 

Comment

 

How to avoid biases in the system?

 

Response

Thank you for the comment. That is a great point raised by the reviewer. Explainability is an interesting way to avoid any bias in the system. It helps to understand why the system makes certain decisions. We can also track back the steps when there is an error. This is very important to ensure transparency in the system. Also, having a user-centric personalized solution is another way to ensure the system is unbiased. This ensures that we take into account the human factor while providing the solution. We updated Section 3.5 with more explanations in this regard. 

 

Comment

 

 

 

For the scenario presented in the implementation section, the processing part in the system is quite simple, it would be helpful if the authors can highlight the usefulness of the AI system. Additionally, it would be better if the authors can present some results in the manuscript. 

Response

Thank you for the comment.  We provide more explanations in the results and conclusion sections to highlight the usefulness of the proposed AI system.

 

Round 2

Reviewer 3 Report

The authors have addressed all of my comments. However, there is just one minor mistake presented in line 303, "Once the vehicles are identified, a personalized dashboard (see Figure ??) is built for.....", the "Figure ??" is referring to "Figure 4"?

Author Response

Point 1: There is just one minor mistake presented in line 303, "Once the vehicles are identified, a personalized dashboard (see Figure ??) is built for.....", the "Figure ??" is referring to "Figure 4"?

Response: Thanks for the comment. Yes, it refers to Figure 4. We corrected the missing figure number. 

Author Response File: Author Response.pdf

Back to TopTop