Effects of Automation and Fatigue on Drivers from Various Age Groups
Round 1
Reviewer 1 Report
The submitted paper describes a driving experiment with just over ninety participants to investigate the effect of automation and fatigue on drivers. The results are analyzed against two genders and two age groups.
Overall the paper is well written and I do not have any obvious or serious concerns about the experimental setup. My comments are as follows:
- The authors should establish the research gap better. Does the literature not cover the research questions identified in section 1.3? And how does this work fill the gap in ways that others haven’t?
- The term “counterbalanced order” is mentioned a lot throughout the paper with no clear definition from the authors, leaving the readers to offer their own interpretation. While the readers’ interpretation is likely to be correct, creating a common understanding by making the wording more specific would ensure the readers get the right message from the paper.
- The figures depicting the statistics for each metric (for example Figure 7, for reaction time) are unconventional and do not make the information clear. I would either choose a table format with two rows for the two conditions (automated and manual) which would make it easy to see the numbers and the differences between them, or show the distribution of data in two overlaid histograms.
Author Response
Dear Reviewer,
Thanks for your comments on our manuscript. We edited the manuscript based on your comments and answered them in the attached Table.
Best regards,
Authors
Author Response File: 
Author Response.docx
Reviewer 2 Report
The authors present a driving simulator study investigating driver fatigue and automation. This is an important area of research, particularly when driver fallback is required. Overall, my main concern with the manuscript is the somewhat vague description of the sample of participants and presentation of results. This makes it a little difficult for the reader to determine how representative the sample of participants were who undertook the study and how meaningful the findings are. If the authors address these issues and some minor comments below I think the manuscript would be suitable for publication.
Minor comments:
Introduction
- The first sentence is vague and is not needed
- Line 30 – remove “were” from sentence
- Line 38 – use “they” instead of he/she
- Line 51 – remove the bracket
- Paragraph starting at line 55. The writing is vague. Include information about what the differences were that were identified in the study.
- Line 59 – introduce acronym for adaptive cruise control when it is first used in the text. Similar comment for other acronyms, ADS, ADGS, PERCLOS, EEG etc.
- Line 69 – what were the significant findings of Muslim? elaborate on the findings from the literature.
- Gaps in the reviewed literature are not highlighted in the introduction. The author needs to provide stronger justification for this study and highlight the contribution to knowledge.
- There is no mention of driver workload research in the introduction, yet it is a key dependent variable of the study.
Materials and methods
- The heading number is incorrect
- Table 1 – include distractive statistics for the whole sample as well as sub-groups by age and gender. Also include the sample size for each subgroup. Without this information it is not clear if the authors have a well-balanced study.
- Line 174 – explain the scoring used for the NASA TLI.
- Table 2 – is incorrectly labeled. Also is this table really needed. It is a 2x2 study design testing rested condition and driving mode. The table makes the study seem more complicated than it really is.
Results
- Present a summary table of findings (mean and SD) for dependent variables tested by age, gender and condition. Then discuss significant findings.
- Figure 2 – include error bars (same comment for other graphs) and report scores for frustration.
Author Response
Dear Reviewer,
Thanks for your comments on our manuscript. We edited the manuscript based on your comments and answered them in the attached Table.
Best regards,
Authors
Author Response File: Author Response.docx
Reviewer 3 Report
The paper has some merits, however, authors need to address the following items before the paper gets accepted.
Introduction and Literature Review:
Please elaborate further of how this study differs from Muslim et al. [21] and Scarpelli et al. [31]
What are the gaps of existing studies and how does this study fill those gaps.
Method:
Can author provide more on the selection of dependent variables (age and gender)
Figure 2 and Figure 3. Is there any difference between young and old drivers’ mental demand workload during automated and manual driving?
Figure 6: can the author explain further why baseline KSS scores of fatigues is a lot higher than baseline KSS score of rested?
In relation to Figure 7. can authors elaborate on what would be reasons why Driving automation slowed down the reaction time (RT) of drivers and how these information can assist automated driving?
Can the author provide justification on the selection of age groups (20-49 years and 50-85 years)
Since gender effects did not reach statistical significance, it is recommended to modify the title and focus the discussion on age only.
Conclusion:
The author can extend the conclusion by elaborating on how the findings of this study differ from existing studies (e.g., Soares et al. [25])
Author Response
Dear Reviewer,
Thanks for your comments on our manuscript. We edited the manuscript based on your comments and answered them in the attached Table.
Best regards,
Authors
Author Response File: Author Response.docx
Reviewer 4 Report
The main purpose of this contribution is to investigate the effects of fatigue and automation on drivers' reaction time, perceived workload, and perceived fatigue. There are already several studies in the same direction, however, the present work has mainly focused on the interrelation of several parameters that vary simultaneously, which is lacking in the current research. Another strength of this work is the higher number of participants that took part in the experiment compared to most other studies in this field, which makes the statistical results more robust.
Conceptual comments:
- How is the threshold between two age groups chosen (why 50)? Drivers around 50 appear in both groups and will be similar, and this will contaminate the data and make it difficult to find the differences between the two age groups.
- In this study, drivers are asked to react to a visual stimulus by pressing the steering wheel. Is this type of reaction comparable with the takeover situation in Level 3 of automation? In driving scenarios, situation awareness plays a significant role which is not considered here.
- How is the appearance time of the coffee cup distributed? If it appears right at the beginning of the drive, for example, the reaction time does not show the actual effect of the automation mode. Also, different appearance times result in different learning effects.
- How is the rested/fatigued condition of drivers measured? There are different levels of fatigue, but here the levels are not considered.
- Different people have different sleeping habits. For some, 4 hours of sleep at night is enough. Have you considered this parameter when distinguishing between rested and fatigued?
- Have you already considered the different effects of extended wakefulness and sleep deprivation? Do they have the same effect on human performance or not? And if they are not the same, what percentage of drivers belong to each of these groups?
- Results section
- According to Table 1, the driving activity/experience of participants is not evenly distributed (between young and older groups and between female and male groups (medians)). Does this affect the results and how?
- Table 3: In the rested condition (in both automated and manual driving modes), subjective and objective fatigue (PERCLOS) are not correlated! How can this be explained?
- For PERCLOS and RT different metrics (maximum, minimum, median, mean, variance) are included. Can you explain what can be inferred from these metrics? For example, if fatigue has a significant correlation with minimum PERCLOS but not with the other PERCLOS metrics, what can be inferred?
- In this study, four independent variables (fatigue, automation, gender, age) and four dependent variables (sub. workload, sub. fatigue, obj. fatigue, reaction time) are examined and in the results section, most of the relations are presented. But still an overview of all the examined variables and their interrelation is missing. Maybe adding another visualization method can improve it. To visualize multivariate data (three or more variables) spider charts and tables can be helpful
Specific points:
- Line 41 ”Therefore, it is not safe to hand over the car’s control to the drowsy drivers in automated cars which can increase the risk of accidents.”: The previous sentences are about manual driving and this sentence cannot be deduced from them because here the situation is about hand over situation which happens during automated mode. There are even some researches, which show that drowsy drivers in takeover situations show better/faster reactions [Kreuzmair et al., The influence of driver fatigue on take-over performance in highly automated vehicles].
- Line 126: When the word "fatigue" is used in the text, it should be clarified whether it is actual fatigue (which occurs in drivers), subjective fatigue (which is rated by drivers), or objective fatigue (which is measured by PERCLOS).
- Line 130: Either section 2 is missing or the numbering of sections is wrong.
- Line 132: Ninety-two
- Line 172: An explanation about the experiment procedure is missing. Which instructions did the drivers receive? Were they instructed to react fast (high criticality) or not? Were they instructed to monitor the driving situation and react if necessary or they were only expected to monitor the coffee cup sign? Was the experiment in the morning, noon, or afternoon? When they filled the questionnaires?
- Table 2: what is d?
- References section
- Ref 12: Provide a translation of the title in square brackets.
- Ref 18: incomplete
- Ref 37: not mentioned in the text
- Ref 40: the publisher is missing
Author Response
Dear Reviewer,
Thanks for your comments on our manuscript. We edited the manuscript based on your comments and answered them in the attached Table.
Best regards,
Authors
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
The authors have addressed my previous comments. I would like the authors to include comment regarding what is shown by the error bars in Figures 2 - 5. Convention would be to include 95% CI. They appear to be SD, but are not labeled, so it is difficult to be sure. Following this change I think the manuscript is suitable for publication.
Author Response
Dear Reviewer,
Thanks for your comments on our manuscript. We edited the manuscript based on them and answered them in the attached Table.
Best regards,
Authors.
Author Response File: Author Response.docx
Reviewer 3 Report
I thank the authors to address all comments, however, the provided justification and reasoning for the grouping of age groups are insufficient, hence greatly appreciated if the authors can elaborate more on that and provide references on this matter.
The authors cited studies on the effects of age on the accident (26, 28, 27 and others) which clearly used different age groups than that used in this study. For example study by Obst et al (26) found that drivers aged between 25 and 34 years to accept the high level of risk associated with operating a vehicle while sleepy, hence it will be better if the authors can relate their age group mechanism with these findings.
by discussing the characteristics .of driver for each group and how AV can play a role in strengthening those characteristic or even lessening the effects of "bad" characteristics.
Author Response
Dear Reviewer,
Thanks for your comments on our manuscript. We edited the paper based on them. They are also answered in the attached Table.
Best regards,
Authors
Author Response File: Author Response.docx
Reviewer 4 Report
Thank you for your effort. The mentioned points are incorporated, the tables provided in supplementary are informative and the graphs in results part are improved. Also, the added comments clarified the parts that were unclear. Some minor points should be edited before publication:
- Title is without the word “Gender” and the citation on the first page includes this word!
- Please add the translation of the German reference [13] to its citation.
Author Response
Dear Reviewer,
Thanks for your comments on our paper. We answered them in the attached Table.
Best regards,
Authors
Author Response File: Author Response.docx
