Impact of Age on Takeover Behavior in Automated Driving in Complex Traffic Situations: A Case Study of Beijing, China

Round 1

Reviewer 1 Report

The topic is relevant and actual. The authors justified the need to deal with the topic about impact of age on takeover of automated driving. The content of the article fills the gap in this area of knowledge.

Author Response

Response to Reviewer 1 Comments

Point 1: Topics about practical implications. Well-described research methodology, precisely selected factors determining the test results, a well-characterized stage approach.

I highly value the inclusion of both: travel costs and travel time in the analysis. I also drew attention to the changes in passenger transport behavior, including their preferences. These factors are important from the passenger's point of view.

I also like the categorization of single and multi-stop junctions and stops, enabling the development of timetables in many variants. Thus, it was possible to provide the basis for various solutions, which will allow transport systems to evolve over the long term.

Nanjing example clearly described. It allowed to verify model assumptions. This allows to recommend the model to other agglomerations.


Response 1: Thanks for the reviewer for affirmation and appreciation of the paper.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments to the authors

The manuscript deals with a key topic for automated driving, that has still not been explored in full detail. Indeed, the identification of older drivers’ peculiarities when driving an automated vehicle represents a key point, not only from a driving safety perspective, but also for the development of the vehicles themselves. Therefore, the contribution of the manuscript to the field may be very important. However, I have some concerns that prevent me from accepting the manuscript without first addressing the following major and minor revisions.

Abstract

• Please, better describe the experimental design and add a line regarding the limitations of the study

Introduction

• I would suggest to better integrate the first part of the Introduction (dealing with the topic of older drivers in general) with the second part (automated driving and takeover issues). Right now, the logic that guided the work is quite implicit and not always well delineated. Maybe the inclusion of more works related to these fields would help.
• Please, add the hypotheses that guided your work at the end of the introduction, stating them explicitly

Methodology

• Please, add a table or a summary of the demographics of the three groups (age, sex, driving experience, etc.)
• Which criterium (or criteria) were used to divide the three groups? 60 years old does not necessarily mean the participant can be considered as older drivers. Please, provide justification.
• Was any a-priori power analysis run to determine the number of participants to be included in the study?
• Was the informed consent collected for all the participants? Please, add it to the text.
• The experimental design is not clearly described, and this makes difficult a clear understanding of the entire work. Apparently, the study consists of a mixed experimental design with 16 within-participants conditions and three groups, but then the statistical analyses were done separately, not taking into account the interaction possibly caused by the within-participants conditions (see also below).
• Table 1: how was the “vehicle in stability” determined? Is it another warning signal? Please, clarify all the definitions and provide justification of how and why these variables were chosen (among the many possible driving variables you may want to use for the same purposes).
• Driving simulator: please, add a picture of the simulator. Why were the participant invited to drive at 100 km/h when the limit was 120 km/h?
• Line 124/125: “The length 123 of the whole experimental section was 1000m (the length in fog zone was 2500m)”: if the experimental section length corresponded to 1000 m, how is it possible having a 2500 m fog zone? What do you mean with experimental section? Does it correspond to the driving scenario?
• Please, include some pictures of the four driving scenarios

Results

• The statistical analyses apparently are not consistent with the experimental design. I understand that the presence of several within-participants conditions in a single analysis would complicate the results understanding and probably would saturate the model. However, I strongly recommend finding an alternative to analyze the data in a more consistent way. Maybe a change in the approach (using non-parametric statistics or using regression analysis) would help. In any case, the main issue is that the number of participants included in the study is probably not enough to reach a good power.
• How is it possible to have 756 takeover events in the dataset? 42 (participants) × 16 (conditions) = 672. Please, clarify.
• Line 161 and along the entire results section: the statistical results should be reported indicating the statistic value (F), degree of freedom, p value and associated effect size (partial eta squared may work). Moreover, you should report between brackets the mean of the variables in the conditions of interest).
• Please, remove the sentences where you interpret the results (e.g., Line 167/168) from these sections: the interpretation of the results should be done in the discussion.
• Table 2: you should not report the results that reached significance with alpha = 0.10. They cannot be considered as significant. Any interpretation of these results should be removed from the manuscript or, at least, clearly indicated as based on speculation due to the marginal significance of the result.
• Table 2: I would suggest removing the last two columns on the right, leaving only the asterisks to guide the reader in the interpretation of the results
• Table 3 and Table 4: see previous point.
• Table 5: What does this table represent? Is it a summary of the p-values identified with the analyses? What does indicate the column “Main line”? Please, consider to edit the table to make it clear what does it summarize.
• Where the “young” and “middle” group ever compared? Did any significant difference emerge in any of the variables? If not, please state it clearly and discuss the absence of a significant difference.

Discussion

• Please, better rephrase and divide the discussion into subsection accordingly to the significant results
• Please, try to include more references to better explain the results in the light of the present existing literature in the field

Finally, I suggest an overall language revision and editing of the manuscript, paying also attention to the way in which the references are reported in the text (e.g., Line 70, Daniele is the first name, not the surname of the author – this error is repeated with other authors along the entire manuscript).

Author Response

Response to Reviewer 2 Comments

Point 1: Please, better describe the experimental design and add a line regarding the limitations of the study.


Response 1: We agree with the reviewer and thank the reviewer for this constructive comment. We have added the limitation of the study in the abstract as follows:

“Researches on the influence of age in various automated driving conditions will contribute to understand driving behavior characteristics and develop specific automated driving system. This study aims to analyze the relationship between age and takeover behavior in automated driving, where 16 test conditions had been taken into consideration, including 2 driving task, 2 warning time and 4 driving scenarios. 42 drivers in Beijing, China in 2020 were recruited to participate in static driving simulator with Level 3 (L3) conditional automation to obtain detailed test information of the recorded takeover time, mean speed and mean lateral offset. ANOVA test was proposed to examine the significance among different age groups and conditions. The results confirmed that the reaction time would increase significantly with age and the driving stability of the older was worse than the young and middle group. It was also indicated that the older could not adapt to the complex task well when driving, due to the limited cognitive driving ability. Additionally, higher urgency of the scenario explained the variance in the takeover quality. According to the obtained influencing mechanism, policy implications for the developing the vehicle automation considering the various driving behavior of drivers were put forward, so as to correctly identify the high-risk driving conditions in different age group. For further research, on-road validation will be necessary in order to check for driving simulation related effects.”

Point 2: I would suggest to better integrate the first part of the Introduction (dealing with the topic of older drivers in general) with the second part (automated driving and takeover issues). Right now, the logic that guided the work is quite implicit and not always well delineated. Maybe the inclusion of more works related to these fields would help.

Response 2: We agree with the reviewer and thank the reviewer for this constructive comment. We have integrated the first part and the second part by a clearer logic and complemented the necessary description of the relationship between the first and the second part as follows:

“Nowadays, China is witnessing a major change in the proportion of older drivers in road traffic. The number of the drivers aged 60 and older have increased 7.5 million in the past five years, and there are more than 15 million older drivers until 2020. It is controversially discussed whether the elderly have increased involvement in accidents. In 2019, the accident rate of the elderly in China was three times higher compared to the drivers aged from 21-25 and 36-50. Several studies have explored that older drivers are more likely to make mistakes in driving and be at fault in a crash, using the method of statistical analysis. In addition, older drivers differ from younger drivers in the type of crashes (1; 2), and show compensatory driving behavior. Therefore, the safety of the older drivers has become a problem that must be addressed by society. While the automating driving task has become an important research field within the last years, drivers can engage in secondary activities while traveling to their destination. It may propose a new way to prompt the safety of the older drivers.  In Level 3 (L3) conditional automation, the driver is required to neither monitor the driving environment or the automated system performance. Nevertheless, the driver still has to be available for taking over vehicle control in “situations that exceed the operational limits of the automated driving system”, detected and announced by the automated system. Therefore, it is essential to analyze the behavior of the older drivers in highly automated driving in varying conditions.

Aging is accompanied by a decline in cognitive functions and since driving is a complex task, such impairments may also be relevant for safe road behavior. Abundant studies have explored whether the driving ability of older drivers will decline as drivers grow older, based on the simulated driving data in regular driving behavior scenarios without autonomous driving scenarios(3-6). Older drivers experienced significantly slower reaction times, had more collisions, drove slower, deviated less in speed, and were less able to maintain a constant distance behind a pace car (7). Depestele analyzed 22 studies and also concluded that older persons had a more variable, less consistent driving simulator performance, such as more variable speed adaptation or less consistent lane keeping behavior (8). Nakano and Wood found that age sensitivity was significantly associated with the driver safety ratings of the AMD drivers (9; 10). Significant age differences in simulated driving performance were demonstrated(11-14). Furthermore, driving tasks and scenarios were also analyzed to identify the age-related driving performance(15-17). Bunce revealed that older drivers exhibited significantly greater performance inconsistency, particularly marked in the faster motorway condition (18). On the dual carriageway, both young and older drivers drove at similar average speeds, while older drivers showed a 50% higher standard deviation (SD) in steering wheel angular velocity regard to steering stability. On the mountain road, older drivers again demonstrated more difficulty in maintaining correct lane position.  Driving through the inner-city circuit, the SD of the steering wheel angular velocity was approximately 17% higher for the older drivers, and the older participants drove significantly slower in this section (19; 20). Rumschlag found that texting skill level and driver age were significantly correlated with the percent of subjects exhibiting Lane Excursions, where the highly text skilled drivers’ age was significantly correlated with the number of Lane Excursions, the percent of subjects exhibiting Lane Excursions and the percent of texting time in Lane Excursions(21).

On the aspect of the automating driving task, a large variety of take-over studies have been conducted within the past years, giving insight into drivers’ behavior in take-over situations, where most of the research have not focused on the influence of the age. Firstly, the training system of automating driving task was examined and Sportillo indicated that the light VR training is preferred with respect to the other systems (22). Then, dependent variables, including reaction time, maximum acceleration, average lane departure distance, number of collisions and so on, were collected and analyzed (23; 24). A shorter mean take-over time was associated with a higher urgency of the situation, not using a handheld device, not performing a visual non-driving task, having experienced another take-over scenario before in the experiment, and receiving an auditory or vibrotactile take-over request, while the mean and standard deviation of the take-over time were highly correlated (25). Gold investigated the effect of the traffic in takeover situations and the 20-Questions task, and the report showed that the presence of traffic led to longer takeover times and worse takeover quality in the form of shorter time to collision and more collisions, while the 20-Questions task did not influence takeover time but seemed to have minor effects on the takeover quality. Furthermore, a few researchers investigated the influence of age on the take-over of vehicle control in highly automated driving. Körber found that older drivers reacted as fast as younger drivers, however, they differed in their modus operandi as they braked more often and more strongly and maintained a higher time-to-collision(26). Goldrevealed that especially the time budget, traffic density and the repetition strongly influenced the take-over performance, while the non-driving related tasks, the lane and drivers’ age explained a minor portion of the variance in the take-over performances (27).”

Point 3: Please, add the hypotheses that guided your work at the end of the introduction, stating them explicitly

Response 3: We agree with the reviewer and thank the reviewer for this constructive comment. The hypotheses have been added at the end of the introduction, including the driving behavior of older drivers differs with the young drivers and varies under different automating driving tasks. It is can be seen as follows:

“Considering the significant increases in number of older drivers and the development of the automating driving technology, existing studies have only focused on identifying the differences in driving ability between young and older drivers in terms of manual driving. However, the age effects existing in the automated driving behavior have not been fully considered or analyzed, including the older drivers’ takeover time, driving stability and so on. Based on the hypotheses that the driving behavior of older drivers differs with the young drivers and varies under different automating driving task, this study records and calculates the data of takeover behavior in automated driving, including the takeover time, mean speed and mean lateral offset. Specially, 16 test conditions have been taken into consideration, including driving task, warning time and driving scenarios.”

Point 4: Please, add a table or a summary of the demographics of the three groups (age, sex, driving experience, etc.)

Response 4: We agree with the reviewer and thank the reviewer for this constructive comment. The statistics of participants personal attributes has been added by the form of a table, which can be seen as follows:

Table 1. Statistics of participants personal attributes

Point 5: Which criterium (or criteria) were used to divide the three groups? 60 years old does not necessarily mean the participant can be considered as older drivers. Please, provide justification.

Response 5: We agree with the reviewer and thank the reviewer for this constructive comment. The principle to classify the group is based on the age classification standard in China. At the same time, we also added the supplementary description in the Methodology as follows:

“42 participants (Mean age=42.53 SD=15.85) were recruited to take part in this experiment and signed the informed consent form, including 10 females and 32 males. Based on the age classification standard in China, the participants were classified into three group, where the young group (n = 15, M = 25.4 years) ranges from 18 to 30 years, the middle group (n = 14, M = 43.3 years) ranges from 31 to 60 years and the old group (n = 13, M = 64.7 years) ranges from 61 to 82 years. The possession of a driving license for at least one year was required for participation and participants held their license for a mean of 17.06 (SD = 12.17) years. (Table 1)”

Point 6: Was any a-priori power analysis run to determine the number of participants to be included in the study?

Response 6: We agree with the reviewer and thank the reviewer for this constructive comment. We have referenced the other similar research papers, where the number of the participants ranges from 29 to 60. Combined the actual recruitment situation, our research finally recruited 42 participants.

Point 7: Was the informed consent collected for all the participants? Please, add it to the text.

Response 7: We agree with the reviewer and thank the reviewer for this constructive comment. The informed consent was collected for all the participants in this study, and we have added it to the text as follows:

“42 participants (Mean age=42.53 SD=15.85) were recruited to take part in this experiment and signed the informed consent form, including 10 females and 32 males. Based on the age classification standard in China, the participants were classified into three group, where the young group (n = 15, M = 25.4 years) ranges from 18 to 30 years, the middle group (n = 14, M = 43.3 years) ranges from 31 to 60 years and the old group (n = 13, M = 64.7 years) ranges from 61 to 82 years. The possession of a driving license for at least one year was required for participation and participants held their license for a mean of 17.06 (SD = 12.17) years. (Table 1)”

Point 8: The experimental design is not clearly described, and this makes difficult a clear understanding of the entire work. Apparently, the study consists of a mixed experimental design with 16 within-participant’s conditions and three groups, but then the statistical analyses were done separately, not taking into account the interaction possibly caused by the within-participants conditions (see also below).

Response 8: We agree with the reviewer and thank the reviewer for this constructive comment. in this study, we mainly focused on revealing whether the age would impact the driving in automating driving task and the older drivers’ driving behaviour would vary under different automating driving task. In order to better describe the experimental design, the part of Study design and measures have been revised as follows:

“Each participant drove around the circuit under four different driving scenarios (Main-line, On-ramp, Fog-cluster and Accident driving scenarios) each for four times, once for each of the two driving tasks (working state and entertainment state driving task) and two warning time (5s or 10s). One participant needed to complete 16 tests in total and the order of tests was random. The relationship between age and driving task, warning time, and driving scenarios was analyzed for this experiment.”

Point 9: Table 1: how was the “vehicle in stability” determined? Is it another warning signal? Please, clarify all the definitions and provide justification of how and why these variables were chosen (among the many possible driving variables you may want to use for the same purposes).

Response 9: We agree with the reviewer and thank the reviewer for this constructive comment. When the speed of the vehicle reaches varies within 1km/h, it is regarded as a stable state. We have also supplement it into the text as follows:

“Dependent variables were take-over time (TOT), control time (CT), mean speed (MS) and mean lateral offset (MLO). Taking over time was defined the time between the warning signal and the takeover, which changed the automated driving mode into manual driving mode. The control time was defined the time between the warning signal and the first input after takeover using brake pedal, steering wheel or accelerator. When the speed of the vehicle reaches varies within 1km/h, it is regarded as a stable state. Take-over quality was measured in this study by mean speed and mean lateral offset. Table 2 summarizes the dependent variables.”

Point 10: Driving simulator: please, add a picture of the simulator. Why was the participant invited to drive at 100 km/h when the limit was 120 km/h?

Response 10: We agree with the reviewer and thank the reviewer for this constructive comment. The free-flow speed was designed for 120km/h on average, which was the average speed of the simulated driving vehicle except the participant. While the participants drove highly automated at a speed of 100km/h on average. We also revised the description of the 2.3 Apparatus as follows:

“Driving simulator: The experiment was conducted in a static driving simulator with Level 3 (L3) conditional automation, composed of steering wheel, throttle, brake and dashboard. Data were recorded at a frequency of 20 Hz and transferred to central control platform. Participants drove highly automated at a speed of 100km/h on average on a four-lane highway, where the free-flow speed was 120km/h and the width of road cross section was 26m (Lane width=3.75m, width of green belt=2m, width of road shoulder=2m). The length of the whole experimental section was 1000m (the length in fog zone was 250m), and the starting point was 200m apart from the starting point of warning time, as shown in Figure 1.”

Point 11: Line 124/125: “The length 123 of the whole experimental section was 1000m (the length in fog zone was 2500m)”: if the experimental section length corresponded to 1000 m, how is it possible having a 2500 m fog zone? What do you mean with experimental section? Does it correspond to the driving scenario?

Response 11: We agree with the reviewer and thank the reviewer for this constructive comment. This is a type mistake, and we have replaced the 2500m with 250m as follows:

“Driving simulator: The experiment was conducted in a static driving simulator with Level 3 (L3) conditional automation, composed of steering wheel, throttle, brake and dashboard. Data were recorded at a frequency of 20 Hz and transferred to central control platform. Participants drove highly automated at a speed of 100km/h on average on a four-lane highway, where the free-flow speed was 120km/h and the width of road cross section was 26m (Lane width=3.75m, width of green belt=2m, width of road shoulder=2m). The length of the whole experimental section was 1000m (the length in fog zone was 250m), and the starting point was 200m apart from the starting point of warning time, as shown in Figure 1.”

Point 12: Please, include some pictures of the four driving scenarios

Response 12: We agree with the reviewer and thank the reviewer for this constructive comment. The simulated driving equipment and environment were specially designed for this study, which was still under in the process of optimization. We would add the related data and pictures in the future study.

Point 13: The statistical analyses apparently are not consistent with the experimental design. I understand that the presence of several within-participants conditions in a single analysis would complicate the results understanding and probably would saturate the model. However, I strongly recommend finding an alternative to analyze the data in a more consistent way. Maybe a change in the approach (using non-parametric statistics or using regression analysis) would help. In any case, the main issue is that the number of participants included in the study is probably not enough to reach a good power.

Response 13: We agree with the reviewer and thank the reviewer for this constructive comment. In order to make the result of this study more intuitive, we have revised all the description structure of the result. We firstly analyse whether the age would influence the driving behaviour in automating driving task, after that the impact of the driving tasks are described, which is consistent with the experimental design and the hypotheses in the Introduction.

Point 14: How is it possible to have 756 takeover events in the dataset? 42 (participants) × 16 (conditions) = 672. Please, clarify.

Response 14: We agree with the reviewer and thank the reviewer for this constructive comment. This is a type mistake, and we have replaced the right data as follows:

“A total of 672 takeover data were obtained in this test, where 2 participants failed to take over the vehicle. Finally, 670 valid data were obtained. We formally tested age difference through 3(young, middle, old) ×2(driving task) Analyses of Variance (ANOVA). Table 3 showed that the main effects for age and driving task between the older drivers were significant in takeover time. The TOT and CT under working state driving task suggested high correlation with the age (pTOT(Young×Old) =0.000, pTOT(Middle×Old) =0.006, pCT(Young×Old) =0.072, pCT(Middle×Old) =0.003). The older age was associated with higher reaction time in TOT and CT, where the older drivers react 0.6s slower than the young and middle group on average. While the TOT under entertainment state driving task of the older drivers was similar to the young group, where the TOTs are all equal to 3.73s. Considering the takeover quality, the ANOVA indicated that the age had no significant impact under working task except that the MLO of the old group was more than 0.04m higher than the young and middle group on avereage. On the aspect of the Entertainment task, the middle and old group showed greater difference in takeover quality (pMS-E (Middle×Old) =0.024, pMLO-W (Middle×Old) =0.002), while the young and old group only showed difference in MLO (pMLO-E (Young×Old) =0.011).”

Point 15: Line 161 and along the entire results section: the statistical results should be reported indicating the statistic value (F), degree of freedom, p value and associated effect size (partial eta squared may work). Moreover, you should report between brackets the mean of the variables in the conditions of interest).

Response 15: We agree with the reviewer and thank the reviewer for this constructive comment. All the statistic value (mean value and standard deviation), degree of freedom and p value were all listed in the Tables of Result. We have analysed the significant of the result in the text. In order to make the result more intuitive, we have rewritten all the result section.

Point 16: Please, remove the sentences where you interpret the results (e.g., Line 167/168) from these sections: the interpretation of the results should be done in the discussion.

Response 16: We agree with the reviewer and thank the reviewer for this constructive comment. We have removed all the interpretations of the results in the Results into the section of Discussion.

Point 17: Table 2: you should not report the results that reached significance with alpha = 0.10. They cannot be considered as significant. Any interpretation of these results should be removed from the manuscript or, at least, clearly indicated as based on speculation due to the marginal significance of the result.

Response 17: We agree with the reviewer and thank the reviewer for this constructive comment. We have remove the * of the result with alpha = 0.10 and related description in the text.

Point 18: Table 2: I would suggest removing the last two columns on the right, leaving only the asterisks to guide the reader in the interpretation of the results

Table 3 and Table 4: see previous point.

Response 18: We thank the reviewer for this constructive comment, please allow us to explain our consideration. In order to guarantee the results more scientific and intuitive, the last two columns on the right can reflect the original results, which may provide reference for the other research in the future study. So we may keep the existing table format, and rewrite all the text part of the result section to better display the outcome.

Point 19: Table 5: What does this table represent? Is it a summary of the p-values identified with the analyses? What does indicate the column “Main line”? Please, consider to edit the table to make it clear what does it summarize.

Response 19: We agree with the reviewer and thank the reviewer for this constructive comment. We have merged the Table 4 and Table 5 into one table, whose structure is the same as Table 2 and Table 3. At the same time, we also revised the description of the results in New Table 5 as follows:

“3.3 Result of different driving scenarios

As shown in Table 5, the TOT of older drivers was 0.5s longer than the young group under mainline and fog-cluster driving scenario (pTOT-Main (Young×Old) =0.072, pTOT-Fog (Young×Old) =0.033). While there is no significant difference of CT between three groups. On the aspect of the driving stablility, the MS of old group was more than 8 km/h higher than the middle group under fog-cluster and accident driving scenario (pMS-Fog (Middle×Old) =0.001, pMS-Accident (Middle×Old) =0.007). At the same time, the MLO strongly increased in the old group (pMLO-Main (Young×Old) =0.02 and pMLO-On-ramp (Middle×Old) =0.001) in the scenario of mainline driving. While the dependent variables of the old group in On-ramp driving scenario were all higher than young and middle group, although the difference was not significant except the MLO between the middle and old group (pMLO-On-ramp (Middle×Old) =0.08).

On the aspect of the driving scenarios, the driving scenarios had no significant impact on the time related to take over. The MS in mainline driving scenario was statistically higher than the other scenarios at a confidence level above 99% regardless the age, where the ranked speed from low to high are On-ramp, Accident, Fog-cluster and Main-lane driving scenario. The inconsistency of MLO revealed that the driving scenario had significant influence on the young and middle group, where all the p-value were smaller than 0.01(pMLO-Young =0.057, pMLO-middle =0.019). The MLO of old group increased 0.12m from 0.49m to 0.61m due to the fog cluster.

Table 5. Results obtained from ANOVA test in different driving scenario

Note: ***, **, * = significance at 1%, 5%, and 10%.”

Point 20: Where the “young” and “middle” group ever compared? Did any significant difference emerge in any of the variables? If not, please state it clearly and discuss the absence of a significant difference.

Response 20: We agree with the reviewer and thank the reviewer for this constructive comment. This study mainly focused on the driving behaviour of the older drivers, so the description of the results was mainly related to the differences between the older group with the young and middle group. In order to make the result more exhaustive, we have added the significant difference in young and middle group in the section of Results and Discussion.

Point 21: Please, better rephrase and divide the discussion into subsection accordingly to the significant results

Response 21: We agree with the reviewer and thank the reviewer for this constructive comment. We have rephrased the and divided the discussion into subsections as follows:

“In this study, we investigated the influence of age, driving task, warning time and driving scenario on the takeover time and takeover quality, where the takeover time was evaluated by the TOT and CT, and the takeover quality was evaluated by the MS and MLO.

Age: The TOT and CT appears to increase with age overall, while the MS and MLO of the older drivers are all higher than the other groups. It is confirmed that the age had a negative influence on the takeover stability, agreeing with previous findings indicting that older drivers differed in their modus operandi as they braked more often and more strongly and maintained a higher time-to-collision(26; 27).  More precisely, Older drivers would increase the TOT particularly under working state drving task, 5s warning time, 10s warning time,  Main-line and og-cluster scenario, CT in entertainment state driving task, MS under entertainment state driving task , 10 warning time, Fog-cluster and Accident scenario, and MLO under working state driving task, entertainment state driving task, 5s warning time, 10s warning time and  Main-line scenario.The increased MS revealed that the older would drive in Emergencies as fast as possible to keep safe instead of reducing the driving speed

Driving task: The confidence intervals of TOT and CT in old group show a high overlap, while no significant difference in young and middle groups was found. When the complexity of the driving task increased, the takeover time of the older would increase nearly 1 second. It is indicated that the older drivers could not finish complex task, like texting and sending the messages, due to the decreased cognitive driving ability. This is consistent with a similar, previous study by Lu, where the longer the video length, the lower the absolute error of the number of placed cars, the total distance error and geometric difference between the placed cars and the actual cars(23). What’s more, the dependent variable of takeover quality was not impacted by the driving task. However, contrary to our findings, Gold found the 20-Questions Task did not influence takeover time but seemed to have minor effects on the takeover quality(28). One possible explanation for the contradictory results could be that two driving tasks with different complexity were used in this study while the previous study only took one driving task into consideration and compared the result between the task group and non-task group.

Warning time: Increased warning time significantly decreased the CT about 0.5 second on average in all aged group, indicating that the longer warning time could prompt the thinking time about the first input after takeover. At the same time, the takeover quality of young group was more sensitive to the warning time, where the MS decreased from 72.02 km/h to 68.04 km/h to guarantee safety and the MLO increased from 0.45m to 0.50m.

Driving scenario: The driving scenario neither influenced the TOT nor CT in all aged group, but led to a strongly differences in takeover quality, where the ranked speed from low to high are On-ramp, Accident, Fog Cluster and Main-lane driving scenario. This is also consisted with Gold’s research that the presence of traffic in takeover situations led to worse takeover quality (28). While Kim and Zhang all confirmed that the reaction times were significantly different between events, where shorter mean take-over time was associated with a higher urgency of the situation(24; 25). The difference research result may independent variable used to separate the driving scenario was traffic density, while this research used different driving environment like Fog cluster, On-ramp way and Driving accident instead of changing the traffic density.”

Point 22: Please, try to include more references to better explain the results in the light of the present existing literature in the field

Response 22: We agree with the reviewer and thank the reviewer for this constructive comment. We have rewritten the content of the discussion part and highlight the analysis of the significant results.

Point 23: Finally, I suggest an overall language revision and editing of the manuscript, paying also attention to the way in which the references are reported in the text (e.g., Line 70, Daniele is the first name, not the surname of the author – this error is repeated with other authors along the entire manuscript).

Response 23: We agree with the reviewer and thank the reviewer for this constructive comment. We have carefully reviewed the paper and rewritten the right name of the authors.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The quality of the manuscript has now considerably improved. However, some minor aspects still need to be addressed. Moreover, the English language still needs some revisions in order to improve clarity and conciseness.

Point 4: Please, report the statistics divided for the three groups in the new table (young, middle, and older participants).

Point 5: Please, report the criteria you employed also in the manuscript.

Point 6: Please, add these considerations also in the manuscript.

Point 15: For what I saw in the present version of the manuscript, the F values, the degrees of freedom, and the effect size of the significant results are still missing. Please, add them in the new version.

Point 17: Thank you for removing the results with alpha = 0.10. However, all the results with alpha > 0.05 should be removed along the entire manuscript. As an alternative, you may want to clearly label them as marginally significant.

Point 22: I was not able to localize the new references the author added. Please, clarify.