Modeling cognitive load effects of conversation between a passenger and driver

Cognitive load from secondary tasks is a source of distraction causing injuries and fatalities on the roadway. The Detection Response Task (DRT) is an international standard for assessing cognitive load on drivers’ attention that can be performed as a secondary task with little to no measurable effect on the primary driving task. We investigated whether decrements in DRT performance were related to the rate of information processing, levels of response caution, or the non-decision processing of drivers. We had pairs of participants take part in the DRT while performing a simulated driving task, manipulated cognitive load via the conversation between driver and passenger, and observed associated slowing in DRT response time. Fits of the single-bound diffusion model indicated that slowing was mediated by an increase in response caution. We propose the novel hypothesis that, rather than the DRT’s sensitivity to cognitive load being a direct result of a loss of information processing capacity to other tasks, it is an indirect result of a general tendency to be more cautious when making responses in more demanding situations.     

Perceptual Load Induces Inattentional Blindness in Drivers

Perceptual load theory states that the level of perceptual load in a task predicts the processing of task‐irrelevant information. High perceptual load has been shown to result in increased inattentional blindness; however, there is little evidence that this extends beyond artificial computer‐based tasks to real‐world behavior. In this study, we adapted a typical load‐blindness paradigm for use in a driving simulator. Forty‐two drivers performed a series of gap perception tasks where they judged if their vehicle could fit between two parked vehicles, with the task imposing either low or high perceptual load. Awareness for an unexpected pedestrian or animal at the side of the road was found to be significantly lower in the high perceptual load condition. This study is the first to demonstrate perceptual load effects on awareness in an applied setting and has important implications for road safety and future applied research on the perceptual load model.Copyright © 2016 John Wiley & Sons, Ltd.     

The role of drivers’ social interactions in their driving behavior: Empirical evidence and implications for car-following and traffic flow

Models for describing the microscopic driving behavior rarely consider the “social effects” on drivers’ driving decisions. However, social effect can be generated due to interactions with surrounding vehicles and affect drivers’ driving behavior, e.g., the interactions result in imitating the behavior of peer drivers. Therefore, social environment and peer influence can impact the drivers’ instantaneous behavior and shift the individuals’ driving state. This study aims to explore empirical evidence for existence of a social effect, i.e., when a fast-moving vehicle passes a subject vehicle, does the driver mimic the behavior of passing vehicle? High-resolution Basic Safety Message data set (N = 151,380,578) from the Safety Pilot Model Deployment program in Ann Arbor, Michigan, is used to explore the issue. The data relates to positions, speeds, and accelerations of 63 host vehicles traveling in connected vehicles with detailed information on surrounding environment at a frequency of 10 Hz. Rigorous random parameter logit models are estimated to capture the heterogeneity among the observations and to explore if the correlates of social effect can vary both positively and negatively. Results show that subject drivers do mimic the behavior of passing vehicles –in 16 percent of passing events (N = 18,099 total passings occurred in freeways), subject vehicle drivers are observed to follow the passing vehicles accelerating. We found that only 1.2 percent of drivers normally sped up (10 km/hr in 10 s) during their trips, when they were not passed by other vehicles. However, if passed by a high speed vehicle the percentage of drivers who sped up is 16.0 percent. The speed change of at least 10 km/hr within 10 s duration is considered as accelerating threshold. Furthermore, the acceleration of subject vehicle is more likely if the speed of subject driver is higher and more surrounding vehicles are present. Interestingly, if the difference with passing vehicle speed is high, the likelihood of subject driver’s acceleration is lower, consistent with expectation that if such differences are too high, the subject driver may be minimally affected. The study provides new evidence that drivers’ social interactions can change traffic flow and implications of the study results are discussed.     

Safety at first sight? – Manual drivers’ experience and driving behavior at first contact with Level 3 vehicles in mixed traffic on the highway

Soon, manual drivers will interact with conditionally automated vehicles (CAVs; SAE Level 3) in a mixed traffic on highways. As of yet, it is largely unclear how manual drivers will perceive and react to this new type of vehicle. In a driving simulator study with N = 51 participants aged 20 to 71 years (22 female), we examined the experience and driving behavior of manual drivers at first contact with Level 3 vehicles in four realistic driving scenarios (highway entry, overtaking, merging, introduction of a speed limit) that Level 3 vehicles may handle alone once their operational domain extends beyond driving in congested traffic. We also investigated the effect of an external marking via a visual external human–machine interface (eHMI), with participants being randomly assigned to one of three experimental groups (none, correct, incorrect marking). Participants experienced each driving scenario four times, twice with a human-driven vehicle (HDV), and twice with a CAV. After each interaction, participants rated perceived driving mode of the target vehicle as well as perceived safety and comfort. Minimum time headways between participants and target vehicles served as an indicator of safety criticality in the interactions. Results showed manual driver can distinguish CAVs from HDVs based on behavioral differences. In all driving scenarios, participants rated interactions with CAVs at least as safe as interactions with HDVs. The driving data analysis showed that manual driver interactions with CAVs were largely uncritical. However, the CAVs’ strict rule-compliance led to short time headways of following manual drivers in some cases. The eHMI used in this study neither affected the subjective ratings of the manual drivers nor their driving behavior in mixed traffic. Thus, the results do not support the use of eHMIs on the highway, at least not for the eHMI design used in this study.     

Driving experience, attentional focusing, and the recall of recently inspected events

How does a driver’s perception of roadway events change with experience? A laboratory study addressed this question by comparing novice and experienced drivers as they watched video recordings taken from a moving vehicle. While watching the recordings, the drivers had their eye movements monitored. When the recording was paused, memory for immediately prior events was tested, and recall performance related to what the viewer had been inspecting. The recordings were taken from a vehicle as it travelled along a series of roads, and questions were asked about other road users and about roadway features. The experiment asked about the relationship between driving experience and attentional capture, and about the recall of events recently seen. What attracted attention were objects of central interest such as other road users appearing close to the camera, and moving objects. When the memory test was administered immediately after a hazardous event had occurred, such as a pedestrian stepping into the path of the camera vehicle, then there was evidence of attentional focussing and reduced availability of details about incidental objects. Recall performance generally reflected the pattern of eye fixations, but viewers did not always recall details about fixated objects, and were sometimes able to recall information about objects that were not fixated. Experienced drivers recalled more of the incidental events than the novices, but they were similar in their recall of central events. This supports the association between driving experience and the extent of the effective perceptual field.     

Obstacle avoidance under automated steering: Impact on driving and gaze behaviours

Within the context of more and more autonomous vehicles, an automatic lateral control device (AS: Automatic Steering) was used to steer the vehicle along the road without drivers’ intervention. The device was not able to detect and avoid obstacles. The experiment aimed to analyse unexpected obstacle avoidance manoeuvres when lateral control was delegated to automation. It was hypothesized that drivers skirting behaviours and eye movement patterns would be modified with automated steering compared with a control situation without automation. Eighteen participants took part in a driving simulator study. Steering behaviours and eye movements were analysed during obstacle avoidance episodes. Compared with driving without automation, skirting around obstacles was found to be less effective when drivers had to return from automatic steering to manual control. Eye movements were modified in the presence of automatic steering, revealing further ahead visual scanning of the driving environment. Resuming manual control is not only a problem of action performance but is also related to the reorganisation of drivers’ visual strategies linked to drivers’ disengagement from the steering task. Assistance designers should pay particular attention to potential changes in drivers’ activity when carrying out development work on highly automated vehicles.     

What makes a good driver on public roads and race tracks? An interview study

Future vehicles may drive automatically in a human-like manner or contain systems that monitor human driving ability. Algorithms of these systems must have knowledge of criteria of good and safe driving behavior with regard to different driving styles. In the current study, interviews were conducted with 30 drivers, including driving instructors, engineers, and race drivers. The participants were asked to describe good driving on public roads and race tracks, and in some questions were supported with video material. The results were interpreted with the help of Endsley’s model of situation awareness. The interviews showed that there were clear differences between what was considered good driving on the race track and good driving on the public road, where for the former, the driver must touch the limit of the vehicle, whereas, for the latter, the limit should be avoided. However, in both cases, a good driver was characterized by self-confidence, lack of stress, and not being aggressive. Furthermore, it was mentioned that the driver’s posture and viewing behavior are essential components of good driving, which affect the driver’s prediction of events and execution of maneuvers. The implications of our findings for the development of automation technology are discussed. In particular, we see potential in driver posture estimation and argue that automated vehicles excel in perception but may have difficulty making predictions.     

Influence of traffic conditions on driver behavior before making a right turn at an intersection: Analysis of driver behavior based on measured data on an actual road

This paper describes an investigation of the influence of the position of a forward vehicle and following vehicle on the onset of driver preparatory behavior before making a right turn at an intersection. Four experimental vehicles with various sensors and a driving recorder system were developed to measure driver behavior before making a right turn at a specific intersection on a public road. The experimental term was eight weeks to collect data on natural driving maneuvers. The relationships between the remaining distances to the center of the intersection when releasing the accelerator pedal, moving the right foot to cover the brake pedal, and activating the turn signal and the relative distances from the forward and following vehicles were analyzed based on the measured data. The time it took to reach the center of the turn and the driving speed when each behavioral event occurred were also evaluated from the viewpoint of the relative position between the driver’s vehicle and the leading or following vehicles. The results suggest that the drivers approached the target intersection in a car-following condition, and that the positions of the front and rear vehicles and the vehicle velocity influence the onset location and timing of releasing the accelerator pedal and covering the brake pedal. Drivers began to decelerate closer to the center of the intersection when they approached the intersection close to a leading or following vehicle at a reduced driving speed. However, these influences were not reflected in the turn signal operation, indicating that drivers intend to make a right turn at a constant location while approaching a target intersection and that intention appears in the turn signal activation. The findings of this observational study imply that the method of providing route guidance instruction, in which the traffic conditions surrounding the driver’s vehicle are taken into consideration, is effective in reducing driver errors in receiving instruction and following the correct route. The results also indicate that measuring and accumulating different behavioral indices based on traffic conditions contribute to determining the criteria for the presentation timing just before reaching the intersection, which can assist drivers in preparing to make a right turn at a usual location. Driver decelerating maneuvers are used while driving without leading or following vehicles and while driving with a lead and/or following vehicle at long range, and driver turn signal operations are used when approaching an intersection under close car-following conditions.     

Safe driving and road-crossing tasks: A particular case of successful transfer of learning

The current study examines whether crucial safe driving skills are associated with safe road-crossing skills as pedestrians. The main research question was whether skills that are acquired from the point of view of a driver are associated with the skills of pedestrians in different platforms or settings. Furthermore, the study examines whether task performance on one platform (driving) primes an operator for task performance on another (road-crossing as a pedestrian) or vice versa. Sixty people took part in this study and completed a demographic questionnaire, a Driving Behavior Questionnaire, a Pedestrian Behavior Scale and two computerized tests – a Hazard Perception Test for Drivers and a Hazard Perception Test for Pedestrians.We found that the better the participants detect hazards on the road as drivers, the better they detect hazards as pedestrians as well, and that most of the participants’ self-reported values regarding their driving and their road-crossing as pedestrians are correlated. The study revealed an association between years of seniority in driving and the number of driving hours per week, and some behavioral variables as pedestrians – meaning that exposure to the road as a driver may be related to safer behavior as a pedestrian.     

Attention all ‘drivers’: You could be to blame,no matter your behaviour or the level of vehicle automation

Vehicle crashes are one of the leading causes of human deaths worldwide, with crashes predominately attributed to failures of human drivers. Whilst increasing vehicle automation is argued to reduce road crashes via decreased driver involvement, automation also raises concerns around driver blame and stakeholder responsibility. This study examines blame for crash scenarios across four different forms of driver distraction behaviours (phone, sleep, work and driving under the influence), and across four levels of vehicle automation (no automation [manual], partially automated, highly automated, fully automated), using a mixed (qualitative and quantitative) methods approach. Participants (n = 205) were randomised into one of the four levels of vehicle automation and were presented with vignette crash scenarios involving a pedestrian being hit by a vehicle. Results revealed that scenarios varying driver behavior at the time of the crash, had no significant impact on participants’ blame attribution or selected course of action. The qualitative analysis revealed that despite semantic distinction between some driver behaviours, drivers were deemed responsible for the crash. As automation increased, attribution of blame towards the driver decreased, but did not disappear. Blame simultaneously increased towards other stakeholders including the manufacturer and the government, as level of automation increased. These findings mirror that of previous research and further highlight the need for legal frameworks for crashes with automated vehicles, irrespective of driver behaviours.     

How does glance behavior influence crash and injury risk? A ‘what-if’ counterfactual simulation using crashes and near-crashes from SHRP2

As naturalistic driving data become increasingly available, new analyses are revealing the significance of drivers’ glance behavior in traffic crashes. Due to the rarity of crashes, even in the largest naturalistic datasets, near-crashes are often included in the analyses and used as surrogates for crashes. However, to date we lack a method to assess the extent to which driver glance behavior influences crash and injury risk across both crashes and near-crashes. This paper presents a novel method for estimating crash and injury risk from off-road glance behavior for crashes and near-crashes alike; this method can also be used to evaluate the safety impact of secondary tasks (such as tuning the radio). We apply a ‘what-if’ (counterfactual) simulation to 37 lead-vehicle crashes and 186 lead-vehicle near-crashes from lead-vehicle scenarios identified in the SHRP2 naturalistic driving data. The simulation combines the kinematics of the two conflicting vehicles with a model of driver glance behavior to estimate two probabilities: (1) that each event becomes a crash, and (2) that each event causes a specific level of injury. The usefulness of the method is demonstrated by comparing the crash and injury risk of normal driving with the risks of driving while performing one of three secondary tasks: the Rockwell radio-tuning task and two hypothetical tasks. Alternative applications of the method and its metrics are also discussed. The method presented in this paper can guide the design of safer driver–vehicle interfaces by showing the best tradeoff between the percent of glances that are on-road, the distribution of off-road glances, and the total task time for different tasks.     

The impact of leaving a voicemail,environment familiarity,and pedestrian predictability on driving behavior

Talking on a cell phone can impair driving performance, but the dynamics of this effect are not fully understood. We examined the effects of leaving a voicemail message on driving when there are critical driving targets to attend to (crosswalks and pedestrians). Participants engaged in an ecologically-valid “voicemail” task while navigating a virtual environment using a driving simulator. We also examined the potential weakening or strengthening of effects of leaving a voicemail message on driving as the familiarity and predictability of critical targets changed. Participants completed four experimental runs through the same driving environment in a driving simulator. There were two crosswalks, one with a pedestrian entering the roadway and one without a pedestrian and the location of the pedestrian was predictable (the same pedestrian consistently used the same crosswalk) for the first three runs and then unpredictable for the fourth. Half of the participants left voicemail messages using a hands-free headset, while the other half drove in silence. Leaving a voicemail message increased steering deviation and velocity. Drivers who were leaving a voicemail message decelerated for pedestrians in the roadway to a similar speed as drivers who were not leaving a voicemail message, but they were delayed in braking. Drivers who were leaving a voicemail message also had worse memory for roadway landmarks. These effects were relatively stable across runs through the same driving environment, suggesting that familiarity and predictability did not impact the effects of leaving a voicemail message while driving. Therefore, leaving a voicemail message leads to poorer driving behavior; faster speed, variable steering, and worse memory for roadway landmarks. Interestingly, although drivers who were leaving a voicemail message were slower to react to local targets, they slowed as much as drivers who were not leaving a voicemail message and familiarity with the driving environment did not impact the effects of leaving a voicemail message on driving.     

Driving on familiar roads: Automaticity and inattention blindness

This paper describes our research into the processes that govern driver attention and behavior in familiar, well-practiced situations. The experiment examined the effects of extended practice on inattention blindness and detection of changes to the driving environment in a high-fidelity driving simulator. Participants were paid to drive a simulated road regularly over 3 months of testing. A range of measures, including detection task performance and driving performance, were collected over the course of 20 sessions. Performance from a yoked Control Group who experienced the same road scenarios in a single session was also measured. The data showed changes in what drivers reported noticing indicative of inattention blindness, and declining ratings of mental demand suggesting that many participants were “driving without awareness”. Extended practice also resulted in increased sensitivity for detecting changes to road features associated with vehicle guidance and improved performance on an embedded vehicle detection task (detection of a specific vehicle type). The data provide new light on a “tandem model” of driver behavior that includes both explicit and implicit processes involved in driving performance. The findings also suggest reasons drivers are most likely to crash at locations very near their homes.     

The impact of auditory continual feedback on take-overs in Level 3 automated vehicles

ObjectiveTo implement auditory continual feedback into the interface design of a Level 3 automated vehicle and to test whether gaze behavior and reaction times of drivers improved in take-over situations.BackgroundWhen required to assume manual control in take-over situations, drivers of Level 3 automated vehicles are less likely than conventional drivers to spot potential hazards, and their reaction time is longer. Therefore, it is crucial that the interface of Level 3 automated vehicles will be designed to improve drivers’ performance in take-over situations.MethodIn two experiments, participants drove a simulated route in a Level 3 automated vehicle for 35 min with one imminent take-over event. Participants’ gaze behavior and performance in an imminent take-over event were monitored under one of three auditory interface designs: (1) Continual feedback. A system that provides verbal driving-related feedback; (2) Persistent feedback. A system that provides verbal driving-related feedback and a persistent beep; and (3) Chatter feedback. A system that provides verbal non-driving-related feedback. Also, there was a control group without feedback.ResultsUnder all three auditory feedback designs, the number of drivers' on-road glances increased compared to no feedback, but none of the designs shortened reaction time to the imminent event.ConclusionIncreasing the number of on-road glances during automated driving does not necessarily improve drivers’ attention to the road and their reaction times during take-overs.ApplicationPossible implications for the effectiveness of auditory continual feedback should be considered when designing interfaces for Level 3 automated vehicles.     

Effect of verbal messages with reminders to communicate driving situations to alter driver behavior in conditional driving automation

In conditionally automated driving, drivers are relieved of steering (hands-off), accelerating, and braking actions as well as of continuous monitoring of driving situations and the system operation status (eyes off). This enables continuously engagement in non-driving-related activities. Managing the allocation of a driver’s attention to the surrounding environment and automation status presents a major challenge in human–machine system design. In this study, we propose a verbal message with a reminder (monitoring request) to divert the driver’s attention from non-driving-related activities to peripheral monitoring under conditionally automated driving. When the system encounters events related to weather, traffic, and road geometry, it provides a verbal message pertaining to the road surroundings (e.g., “It is foggy outside”) to the driver. After three seconds, the system provides a reminder message (i.e., “Did you confirm it?”) to the driver. We explore two questions: (1) how does the message with the reminder affect the driver’s attention allocation, and (2) how does the message with the reminder affect the driver behavior in response to a request to intervene (RTI). With a driving simulator, we designed a repeated measures mixed design with a between-participant factor of “Driving condition” and within-participant factors of “Event type” and “Measurement time”. Three driving conditions were established as follows: no messages, messages without reminders, and messages with reminders. Twenty-seven drivers participated as participants in the driving simulator experiment. Results showed that the reminder message was effective in allocating the participants’ attention to the surrounding environment, and the participants took over the driving task after spending more time understanding the take-over situation in the condition of messages with reminders compared to those in the condition of no messages. We conclude that the proposed reminder message can direct drivers’ attention to the road surroundings during conditionally automated driving. In the future, we plan to design adaptive verbal monitoring requests to adjust the reminder message according to the situation.     

Understanding driver anger and aggression: attributional theory in the driving environment

Two studies tested the applicability of Weiner's (1995, 1996, 2001, 2006) attributional model of social conduct to roadway environments. This model highlights the role of inferences of responsibility after making causal judgments for social transgressions. Study 1 employed written scenarios where participants were asked to imagine themselves driving on a major highway. The degree of controllability and intentionality of the driving act was manipulated experimentally by altering the specific event-related details provided to the participants. Study 2 extended this research to life events by having participants complete online driving diaries every 2 days, identifying their most negative/upsetting encounter with another motorist. The most anger-provoking event was selected from among 4 diary entries and participants were asked to respond to a questionnaire similar to that used in Study 1. Path analyses in both studies generally supported predictions derived from Weiner's model; the association between perceived controllability, intentionality, and dispositional locus of causality of the negative driving event and subsequent anger was mediated by perceptions of responsibility. Additional results in Study 2 suggested that low perceived controllability, intentionality, and dispositional locus of causality were associated with reduced perceived responsibility, which, in turn, facilitated feelings of sympathy. Anger was associated with aggressive responses to the offending driver, whereas sympathy was associated with prosocial responses. Recommendations were offered for improved driver safety, including the development of attributional retraining programs to combat self-serving attributional biases, teaching novice drivers about both formal and informal roadway communication, and the promotion of forgiveness among drivers     

Comparison of pedestrian behaviors between drivers and non-drivers in Chinese sample

Road safety is a serious problem worldwide. Pedestrians, as the most vulnerable road users, deserve more attention. The aims of this study were to examine the validity of the Chinese version of the pedestrian behavior scale (CPBS) in both driver and non-driver samples, and to compare pedestrian behaviors between the two samples. In addition, we assessed the association of attention with pedestrian behaviors by exploring the relationships among CPBS, Mindful Attention Awareness Scale (MAAS) and Attention-Related Cognitive Errors Scale (ARCES). Two groups were assessed, including 302 members in the population with driving experience and 307 individuals in the non-driver group without driving experience. All participants completed the CPBS, MAAS, and ARCES, and provided sociodemographic parameters. The results showed that the CPBS had acceptable internal consistency and stability structure. More importantly, pedestrian behaviors were significantly different between drivers and non-drivers. Drivers reported significantly less transgressive and aggressive behaviors compared with non-drivers. As for the relationship between attention and pedestrian behavior, the MAAS score showed a significant negative correlation with aggressive behavior in the CPBS among drivers, while the ARCES score had significant positive correlations with all three CPBS factors. In non-drivers, the MAAS score was negatively correlated with aggressive behavior and positively associated with positive behavior; the ARCES score was positively correlated with aggressive behavior.     

Driver trust in and training for advanced driver assistance systems in Real-World driving

Two-hundred and twenty-three participants completed an online survey regarding their experiences with advanced driver assistance systems (ADAS) on their personal vehicles, with focus on 1) drivers’ trust in 13 ADAS technologies, and 2) perceived effectiveness of currently used methods of training. Eighteen drivers participated in focus groups designed to probe more deeply into survey responses. Results of the survey showed that participant ratings of trust increased significantly with longer vehicle ownership, but participants who experienced unexpected ADAS technology behavior rated their trust over time significantly lower on ADAS technologies with the exception of rear collision avoidance. The majority (75.8%) of participants reported receiving some ADAS instruction at their vehicle dealership, but only 16.6% indicated it was formal. Participants who received formalized training reported it to be significantly more effective than those who received informal overviews of their systems. Use of trial and error and the owner’s manual were the most frequently reported methods of learning outside of dealership training. Responses indicated that the lack of content tailored to trim-specific vehicle features in owner’s manuals was a barrier to effective use.     

Do drivers accept cooperative behavior of their automated vehicle on highways?

Automated cooperatively interacting vehicles will change the future of traffic completely. Such vehicles will be capable of planning and carrying out maneuvers based on vehicle-to-vehicle and vehicle-to-infrastructure communication, enabling a safer driving experience. However, this gain of safety will only be effective if drivers use and accept the decisions made by advanced automated technology. Especially when drivers are cognitively distracted, new strategies might be necessary, e.g., by further explaining the reason for a cooperative decision.In a driving simulator study, we investigated the acceptance of lane change maneuvers in cooperative situations carried out by an automated vehicle on a two-lane German highway. When the automated system detected a potential lane change ahead, it carried out one of three possible maneuvers: accelerate, decelerate, or maintain speed. Participants (N = 20) were asked whether they accepted the system’s behavior either while being cognitively distracted or in an attentive state. Thus, we investigated the influence of a cognitively demanding secondary task and, in addition, further situational characteristics (Scope of action, Criticality for the lane-changing vehicle, Display of intention) on the acceptance towards the system’s behavior. Moreover, participants had to rate the perceived situation’s criticality.Results underlined the importance of explicit indication of the intention to change lanes. Furthermore, increased cognitive load led to a higher acceptance of the defensive system behavior. This study contributes to the development of a user-centered interface and interaction strategy for cooperative interacting vehicles.     

Assessing the effect of long-automated driving operation,repeated take-over requests,and driver’s characteristics on commercial motor vehicle drivers’ driving behavior and reaction time in highly automated vehicles

Automated Commercial Motor Vehicles (CMVs) have the potential to reduce the occurrence of crashes, enhance traffic flow, and reduce the stress of driving to a larger extent. Since fully automated driving (SAE Level 5) is not yet available, automated driving systems cannot perform all driving tasks under all road conditions. Drivers need to regain the vehicle’s control when the system reaches its maximum operational capabilities. This transition from automated to manual is referred to as Take-Over Request (TOR). Evaluating driver’s performance after TORs and assessing effective parameters have gained much attention in recent years. However, few studies have addressed CMV drivers’ driving behavior after TOR and the effect of long-automated driving and repeated TORs. This paper aims to address this gap and gain behavioral insights into CMV drivers’ driving behavior after TOR and assess the effect of the duration of automated operation before TOR, repeated TORs, and driver characteristics (e.g., age, gender, education, and driving history). To accomplish this, we designed a 40-minutes experiment on a driving simulator and assessed the responses of certified CMV drivers to TORs. Drivers’ reaction time and driving behavior indices (e.g., acceleration, velocity, and headway) are compared to continuous manual driving to measure driving behavior differences. Results showed that CMV drivers’ driving behavior changes significantly after the transition to manual regardless of the number of TORs and the duration of automated driving. Findings suggest that 30 min of automated operation intensifies the effect of TOR on driving behaviors. In addition, repeated TOR improves reaction times to TOR and reduces drivers' maximum and minimum speed after TORs. Driver’s age and driving history showed significant effects on reaction time and some driving behavior indices. The findings of this paper provide valuable information to automotive companies and transportation planners on the nature of driver behavior changes due to the carryover effects of manual driving right after automated driving episodes in highly automated vehicles.