Project ARES: Driverless Transportation System. Challenges and Approaches in an Unstructured Road
, David Yagüe 3, Francisco Royo 4, Miguel Ángel de Miguel 3, Francisco Miguel Moreno 3, Alejandro Ruiz-de-la-Cuadra 3, Fernando Viadero-Monasterio 1, Javier Garcia 4, José Luis San Roman 1,2 and José María Armingol 2,3,*Round 1
Reviewer 1 Report
This paper shows a driverless transportation system from project ARES. It is meaningful, but some comments are listed:
1) The contributions of this paper should be concisely summarized in the introduction, and the form of list by point to point is recommended.
2) More details of literature review should be added. For example, the model predictive control is a popular method, such as, “10.1109/TTE.2020.2993862”, “10.1109/TVT.2020.2980169”.
3) How to consider the vehicle stability control?
4) As a whole, this paper is more like a report or magazine rather than a literature paper, since the academics seem to be poor. The related points listed in this paper are too much, and the novelties need to be modified. The result validations are also lacked.
Author Response
Response to Reviewer 1 Comments
Thank you very much for review this work and take time giving us the constructive corrections. We work hard to modify and correct the points you described. Those are the actions taken in the paper in order to improve the quality of the document.
Point 1: The contributions of this paper should be concisely summarized in the introduction, and the form of list by point to point is recommended.
Response 1: It has been added point by point in the introduction the contribution of the paper.
Point 2: More details of literature review should be added. For example, the model predictive control is a popular method, such as, “10.1109/TTE.2020.2993862”, “10.1109/TVT.2020.2980169”.
Response 2: Both sugestion has been added to the introduction of the MPC method.
Point 3: How to consider the vehicle stability control?
Response 3: The stability control for low speed vehicles (less than 20km/h) is not necessary. Regardless, one initiative has been developed where the change between speed references is never bigger than 0.2 m/s² for acceleration and never under -0.3 m/s² for deceleration.
Point 4: As a whole, this paper is more like a report or magazine rather than a literature paper, since the academics seem to be poor. The related points listed in this paper are too much, and the novelties need to be modified. The result validations are also lacked.
Response 4: The global objective of this paper is to show the accepted methods for solving each problem regarding to autonomous vehicles in non conventional road, the aproaches followed and the huge amount of requirements to develop and integrate for the whole architecture. Some more results regarding the control has been added as well as some captures of the testing grounds.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The paper gives a contribution regarding a driverless transportation vehicle intended to work in tourism areas, airports, business centres, testing grounds and similar places with or without mixed traffic. The article is well structured although section 3 could be divided as it is too long.
The main concern is that is stated "Moreover, all the systems have been tested in different track tests with narrow corridors and sinusoidal roads with heavy slopes, in both, simulated and real environments" but no results are presented (only some data when presenting the sensor modules - IMU, LiDAR, etc.). It would be interesting to put results from controllers (MPC, RRC and especially from Speed Manager). Obstacle Manager tests could also be presented.
If possible, results should have been shown in a real environment (with photos and, if possible, with video as supplementary material).
It also would be interesting that the authors answer the following questions and accept the suggestions to improve the paper quality:
- In Figure 2, it would be important to put in the caption the acronyms present in the figure. Although they are explained in the text, placing them in the caption of the figure would make the understanding of the figure clearer.
Author Response
Response to Reviewer 2 Comments
Thank you very much for review this work and take the time to give us constructive corrections. We work hard to modify and correct the points you described. Those are the actions taken in the paper in order to improve the quality of the document.
Point 1: The main concern is that is stated "Moreover, all the systems have been tested in different track tests with narrow corridors and sinusoidal roads with heavy slopes, in both, simulated and real environments" but no results are presented (only some data when presenting the sensor modules - IMU, LiDAR, etc.). It would be interesting to put results from controllers (MPC, RRC and especially from Speed Manager). Obstacle Manager tests could also be presented.
Response 1: Several results has been added to the control section in order to clarify the differences between the methods used. For the obstacle manager, the results in figure 13 describe the detection of the obstalce and emergency stop in 13.b and the detection but not emergency stop in figure 13.c due to the trapezoidal shape.
Point 2: If possible, results should have been shown in a real environment (with photos and, if possible, with video as supplementary material).
Response 2: We add some pictures of the environment at the end of the document.
Point 3: In Figure 2, it would be important to put in the caption the acronyms present in the figure. Although they are explained in the text, placing them in the caption of the figure would make the understanding of the figure clearer.
Response 3: Done. Figure 2 now has all the acronyms listed and described in the caption.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Journal: Electronics (ISSN 2079-9292)
Manuscript ID: electronics-1275735
Title: Project ARES: Driverless Transportation System. Challenges And Approaches In An Unstructured Road
Authors: Pablo Marin-Plaza *, David Yagüe, Francisco Royo, Miguel Ángel de Miguel, Francisco Miguel Moreno, Alejandro Ruiz-de-la-Cuadra, Fernando Viadero-Monasterio, Javier Garcia, José Luis San Roman, José María ç Armingol.
This paper proposed solutions for autonomous driving in rough environments, high slopes, and unstructured terrain without traffic rules. Due to some flaws, this paper needs a major revision. The comments are as follows:
- Although the authors want to propose solutions for autonomous driving in unstructured environments, the innovations of this paper are not clear. The methods mentioned in this paper are some state-of-art algorithms without improvement.
- The symbols in Fig. 6 are not explained.
- In Sec. 3.4.2, 3.5.2, and 3.6.1, the details of solutions should be presented.
- As shown in Sec. 3.6.4, a control manager is used in the software architecture. How about the robustness of this manager by only using the function (6) in Sec. 3.1.2?
Author Response
Response to Reviewer 1 Comments
Point 1: Although the authors want to propose solutions for autonomous driving in unstructured environments, the innovations of this paper are not clear. The methods mentioned in this paper are some state-of-art algorithms without improvement.
Response 1: The intended paper is a gathering of solutions for intelligent transportation systems based on two forms of navigation. One based on localization GPS and control MPC and the other one is based on reactive road controller which is based on perception when localization is not precise enough. Some of the projects described in the introduction are always ready for structured environments and this project work outside conventional roads, hence, the problems are different and the approach to solving them is a singular variation where the novelty is exposed (using managers and the switch between controllers)
Point 2: The symbols in Fig. 6 are not explained.
Response 2: A brief explanation has been added to the document in the corresponding part:
“(represented as a, b, c, etc.) The square shapes in red represent the distance between the nodes and the circular green shape is the search radius from the starting node.”
Point 3: In Sec. 3.4.2, 3.5.2, and 3.6.1, the details of solutions should be presented.
Response 3: All the sections have been improved with more details. Please, give us your opinion on this new version.
Point 4: As shown in Sec. 3.6.4, a control manager is used in the software architecture. How about the robustness of this manager by only using the function (6) in Sec. 3.1.2?
Response 4: The change of the control is based exactly on the formula in Sec 3.1.2 which is the sigmoid function of the covariance from the Kalman filter unscented. This is the primary resource for navigation. If the score is low, there is a change when the reactive road control is taken the action. Before sending the control commands to the vehicle, there is another check for the road detection score. If this score is still low, the vehicle is unable to move and it stops waiting for a good score firstly in localization and secondly in road detection. The robustness is really accurate because the change between modes is almost instantly when the covariance in localization start growing meaning we have lack of 4G for the corrections or GPS signal.
Author Response File: Author Response.pdf
Reviewer 2 Report
The authors have satisfactorily answered all my questions.
Author Response
We did some changes related to the methods in sections 3.4.2, 3.5.2, and 3.6.1, and added a few more images.
Please, check again the result and tell us your opinion.
Best regards.
Round 3
Reviewer 1 Report
Journal: Electronics (ISSN 2079-9292)
Manuscript ID: electronics-1275735
Title: Project ARES: Driverless Transportation System. Challenges And Approaches In An Unstructured Road
Authors: Pablo Marin-Plaza *, David Yagüe, Francisco Royo, Miguel Ángel de Miguel, Francisco Miguel Moreno, Alejandro Ruiz-de-la-Cuadra, Fernando Viadero-Monasterio, Javier Garcia, José Luis San Roman, José María ç Armingol.
This paper proposed solutions for autonomous driving in rough environments, high slopes, and unstructured terrain without traffic rules. The authors have made improvements according to the last review. However, due to some flaws, this paper can be accepted after minor revision. The comments are as follows:
- The eq. 12 is not complete and the constraints in MPC functions are not presented.
- In eq. 8, 9, and 10, the authors should use Greek letters instead of English words.
Author Response
The eq. 12 is not complete and the constraints in MPC functions are not presented.
Now it is complete.
In eq. 8, 9, and 10, the authors should use Greek letters instead of English words.
It is done. We change the comprehensive language in order to make it clear. Please check the text if it is good enough.
Thank you so much for your time and all your efforts.
