Combining a Universal OBD-II Module with Deep Learning to Develop an Eco-Driving Analysis System
Round 1
Reviewer 1 Report
Combining a Universal OBD-II Module with Deep Learning to Develop an Eco-Driving Analysis System
Global comments
Authors develop an OBD-II module which can collect driving information from various car models including fuel consumption. This device can help drivers to make eco-driving.
The manuscript is very well written and structured, and I think it is of interest to readers, so it could be published if some necessary improvements are made.
Specific comments
Figure 7. Some acronyms are not well defined, i.e, AFR, lgt and Thp.
Section 2.6 is very brief and describes a very important part of the research, so its content should be improved and expanded.
The content of figure 8 is not displayed well. It should indicate the scale, the north and highlight the start and end points of the routes. On the other hand, indicate the closest populations to be able to locate the test area. Total length, slope, ADT could be of interest as well.
Figure 9. Instant fuel consumption seems misrepresented. Consumption values above 100 L / 100 km are reflected, and they are values totally out of reality for the three vehicles that have been tested. Mazda 3 around 6 L/100 km, Mitsubishi Lancer 1.8 around 8L/100 km and Toyota Vios around 5L/100 km. In discussion section (line 394) authors confirm instant fuel consumption range 5-110 L/100 km but these are values out of reality. Authors should check the test made and values obtained for the tree cars, because those values are totally out of range (despite being a mountainous area).
Line 311. Replace “Mitsubisui” with “Mitsubishi”
The formats of the titles of the figures and tables are not those of the MDPI template (Palatino Lynotype, font size 9). Please correct them.
Table 4 with the results of “RMSE L/km” and “g” should be changed once consumption values are verified and corrected if necessary.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The authors present a novel idea for enabling an "Eco-Driving Analysis System" by a deep learning approach with OBD-II module. In section 1, the authors start right away with a relatively long literature review, without clarifying the aim and structure of the manuscript. Here, I see a lot of potential for improvement. In the last lines of the introduction is an attempt for this, but it rather not enough content. There is a logical "jump" between section 1 and 2: from system to study. Maybe the authors might think about introducing a section in between called "State-of-the-Art" or "Related Research" with parts of the content of section 1. Section 2 is very descriptive and easy to follow, besides the typical Google-Maps disadvantage of having routes near rivers set in a similar blue; at first sight you see a river and then only understand that it is a calculated shortest path btw. 2 points. Maybe changing the route color would help here. The following results section is as well descriptive, exept the small font size of figures 9 and 10: unfortunately it was impossible for me to read the labels and the elements of the GUI. Conclusions are short, but suitable.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Round 2
Reviewer 1 Report
The authors have improved the document following the proposed recommendations. The most important deficiency of this manuscript was the instantaneous consumptions obtained (above 100 L / 100 km) but I think that this section has been improved and justified the high value of these consumptions with explanations and new tables.
To be published, I only suggest indicating the names of the cities near the experiment carried out (figure 8) that are currently in Chinese in the figure and should be in English to be understood
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
