The Impact of Line-of-Sight and Connected Vehicle Technology on Mitigating and Preventing Crash and Near-Crash Events.

Line-of-sight (LOS) sensors developed in newer vehicles have the potential to help avoid crash and near-crash scenarios with advanced driving-assistance systems; furthermore, connected vehicle technologies (CVT) also have a promising role in advancing vehicle safety. This study used crash and near-crash events from the Second Strategic Highway Research Program Naturalistic Driving Study (SHRP2 NDS) to reconstruct crash events so that the applicable benefit of sensors in LOS systems and CVT can be compared. The benefits of CVT over LOS systems include additional reaction time before a predicted crash, as well as a lower deceleration value needed to prevent a crash. This work acts as a baseline effort to determine the potential safety benefits of CVT-enabled systems over LOS sensors alone.

A Method to Develop the Driver-Adaptive Lane-Keeping Assistance System Based on Real Driver Preferences.

To satisfy the preference of each driver, the development of a Lane-Keeping Assistance (LKA) system that can adapt to individual drivers has become a research hotspot in recent years. However, existing studies have mostly relied on the assumption that the LKA characteristic aligned with the driver's preference is consistent with this driver's naturalistic driving characteristic. Nevertheless, this assumption may not always hold true, causing limitations to the effectiveness of this method. This paper proposes a novel method for a Driver-Adaptive Lane-Keeping Assistance (DALKA) system based on drivers' real preferences. First, metrics are extracted from collected naturalistic driving data using action point theory to describe drivers' naturalistic driving characteristics. Then, the subjective and objective evaluation method is introduced to obtain the real preference of each test driver for the LKA system. Finally, machine learning methods are employed to train a model that relates naturalistic driving characteristics to the drivers' real preferences, and the model-predicted preferences are integrated into the DALKA system. The developed DALKA system is then subjectively evaluated by the drivers. The results show that our DALKA system, developed using this method, can enhance or maintain the subjective evaluations of the LKA system for most drivers.

Effects of lane-change scenarios on lane-change decision and eye movement.

Improper lane-change manoeuvre can cause traffic safety issues and even lead to serious traffic collisions. Quantifying the decision behaviour and eye movements can provide a deeper understanding of lane-change manoeuvre in vehicle interaction environment. The purpose of this study was to investigate the effect of lane-change scenarios defined by gaps on lane-change decision and eye movements. Twenty-eight participants were recruited to complete a naturalistic driving experiment. Eye movements and lane-change decision duration (LDD) were recorded and analysed. Results suggested that the scanning frequency (SF) and saccade duration (SD) were the sensitive parameters to respond to lane-change scenarios. LDD was significantly affected by the scenario, SF, and SD. The increase in LDD was related to the high difficulty gap and high frequency scanning of multiple regions. These findings evaluated the driver's decision performance in response to different lane-change environments and provided valuable information for measuring the driver's scenario perception ability.Practitioner summary: A naturalistic driving experiment was conducted to evaluate the interaction of lane-change decision, eye movement, and lane changing gap in a lane-change task. The results reveal the sensitive eye movement parameters to lane-change scenario, which provide guidelines for driver's perception ability test and professional driver assessment.

Event-related driver stress detection with smartphones in an urban environment: a naturalistic driving study.

Driving in urban areas can be challenging and encounter acute stress. To detect driver stress, collecting data on real roads without interfering the driver is preferred. A smartphone-based data collection protocol was developed to support a naturalistic driving study. Sixty-one participants drove on predetermined real road routes, and driving information as well as physiological, psychological, and facial data were collected. The algorithm identified potentially stressful events based on the collected data. Participants classified these events as low, medium, or highly stressful events by watching recorded videos after the experiment. These events were then used to train prediction models. The best model achieved an accuracy of 92.5% in classifying low/medium/highly stressful events. The contribution of physiological, psychological, and facial expression indices and individual profile information was evaluated. The method can be applied to visualise the geographical distribution of stressors, monitor driver behaviour, and help drivers regulate their driving habits.

The data collection protocol for driving on real roads and the stressful event identification method could potentially be applied for in-vehicle driver status monitoring and stress intervention.

Influence of emotions on the aggressive driving behavior of online -car-hailing drivers based on association rule mining.

Emotion is an important factor that can lead to the occurrence of aggressive driving. This paper proposes an association rule mining-based method for analysing contributing factors associated with aggressive driving behaviour among online car-hailing drivers. We collected drivers' emotion data in real time in a natural driving setting. The findings show that 29 of the top 50 association rules for aggressive driving are related to emotions, revealing a strong relationship between driver emotions and aggressive driving behaviour. The emotions of anger, surprised, happy and disgusted are frequently associated with aggressive driving behaviour. Negative emotions combined with other factors (for example, driving at high speeds and high acceleration rates and with no passengers in the vehicle) are more likely to lead to aggressive driving behaviour than negative emotions alone. The results of this study provide practical implications for the supervision and training of car-hailing drivers.

Based on the association rule mining method, we found a close connection between drivers' emotional states and the manifestation of aggressive driving behaviours. The findings indicate that the combination of negative emotions and various contributing factors significantly amplifies the likelihood of aggressive driving.

Validating self-reported driving behaviours as determinants of real-world driving speeds.

Self-reported driver behaviour has long been a tool used by road safety researchers to classify drivers and to evaluate the impact of interventions yet the relationship with real-world driving is challenging to validate due to the need for extensive, detailed observations of normal driving. This study examines this association by applying the large UDRIVE naturalistic driving study data involving 96 car drivers, comprising 131,462 trips and 1,459,110 km travelled over a duration of 32,096 hours, to compare individual questions and composite indicators based on the Driver Behaviour Questionnaire with real world driving. Self-reported speed behaviour was compared to the measured values under urban and highway conditions. Generalised Linear Mixed Models were developed to examine the relationships between the observed speed behaviours with DBQ errors and violations scores in conjunction with traffic and environmental factors. Drivers' self-reported data on speed selection seldom aligned with their real-world behaviour and there were no meaningful differences between many of the response categories. The DBQ violations and errors scales showed a highly significant correlation with driving speed indicators however they had a low explanatory power compared to other traffic situational and driving factors. Overall, the study highlights the need to validate self-reported driving data against the accuracy and relevance to real-world driving.

Self-reports of driving behaviour have long been a tool in road safety research and evaluation yet responses on speed selection are commonly inaccurate and may have little relation with real-world driving.

Optimal Duration of In-Vehicle Data Recorder Monitoring to Assess Bus Driver Behavior.

This study examined the optimal sampling durations for in-vehicle data recorder (IVDR) data analysis, focusing on professional bus drivers. Vision-based technology (VBT) from Mobileye Inc. is an emerging technology for monitoring driver behavior and enhancing safety in advanced driver assistance systems (ADASs) and autonomous driving. VBT detects hazardous driving events by assessing distances to vehicles. This naturalistic study of 77 male bus drivers aimed to determine the optimal duration for monitoring professional bus driving patterns and the stabilization point in risky driving events over time using VBT and G-sensor-equipped buses. Of the initial cohort, 61 drivers' VBT data and 66 drivers' G-sensor data were suitable for analysis. Findings indicated that achieving a stable driving pattern required approximately 130 h of VBT data and 170 h of G-sensor data with an expected 10% error rate. Deviating downward from these durations led to higher error rates or unreliable data. The study found that VBT and G-sensor data are both valuable tools for driving assessment. Moreover, it underscored the effective application of VBT technology in driving behavior analysis as a way of assessing interventions and refining autonomous vehicle algorithms. These results provide practical recommendations for IVDR researchers, stressing the importance of adequate monitoring durations for reliable and accurate outcomes.

European NCAP Driver State Monitoring Protocols: Prevalence of Distraction in Naturalistic Driving.

OBJECTIVE: examine the prevalence of driver distraction in naturalistic driving when implementing European New Car Assessment Program (Euro NCAP)-defined distraction behaviours. BACKGROUND: The 2023 introduction of Occupant Status monitoring (OSM) into Euro NCAP will accelerate uptake of Driver State Monitoring (DSM). Euro NCAP outlines distraction behaviours that DSM must detect to earn maximum safety points. Distraction behaviour prevalence and driver alerting and intervention frequency have yet to be examined in naturalistic driving. METHOD: Twenty healthcare workers were provided with an instrumented vehicle for approximately two weeks. Data were continuously monitored with automotive grade DSM during daily work commutes, resulting in 168.8 hours of driver head, eye and gaze tracking. RESULTS: Single long distraction events were the most prevalent, with .89 events/hour. Implementing different thresholds for driving-related and driving-unrelated glance regions impacts alerting rates. Lizard glances (primarily gaze movement) occurred more frequently than owl glances (primarily head movement). Visual time-sharing events occurred at a rate of .21 events/hour. CONCLUSION: Euro NCAP-described driver distraction occurs naturalistically. Lizard glances, requiring gaze tracking, occurred in high frequency relative to owl glances, which only require head tracking, indicating that less sophisticated DSM will miss a substantial amount of distraction events. APPLICATION: This work informs OEMs, DSM manufacturers and regulators of the expected alerting rate of Euro NCAP defined distraction behaviours. Alerting rates will vary with protocol implementation, technology capability, and HMI strategies adopted by the OEMs, in turn impacting safety outcomes, user experience and acceptance of DSM technology.

How Do Human-Driven Vehicles Avoid Pedestrians in Interactive Environments? A Naturalistic Driving Study.

One of the major challenges for autonomous vehicles (AVs) is how to drive in shared pedestrian environments. AVs cannot make their decisions and behaviour human-like or natural when they encounter pedestrians with different crossing intentions. The main reasons for this are the lack of natural driving data and the unclear rationale of the human-driven vehicle and pedestrian interaction. This paper aims to understand the underlying behaviour mechanisms using data of pedestrian-vehicle interactions from a naturalistic driving study (NDS). A naturalistic driving test platform was established to collect motion data of human-driven vehicles and pedestrians. A manual pedestrian intention judgment system was first developed to judge the pedestrian crossing intention at every moment in the interaction process. A total of 98 single pedestrian crossing events of interest were screened from 1274 pedestrian-vehicle interaction events under naturalistic driving conditions. Several performance metrics with quantitative data, including TTC, subjective judgment on pedestrian crossing intention (SJPCI), pedestrian position and crossing direction, and vehicle speed and deceleration were analyzed and applied to evaluate human-driven vehicles' yielding behaviour towards pedestrians. The results show how vehicles avoid pedestrians in different interaction scenarios, which are classified based on vehicle deceleration. The behaviour and intention results are needed by future AVs, to enable AVs to avoid pedestrians more naturally, safely, and smoothly.

Does agreement mean accuracy? Evaluating glance annotation in naturalistic driving data.

Naturalistic driving studies often make use of cameras to monitor driver behavior. To analyze the resulting video images, human annotation is often adopted. These annotations then serve as the 'gold standard' to train and evaluate automated computer vision algorithms, even though it is uncertain how accurate human annotation is. In this study, we provide a first evaluation of glance direction annotation by comparing instructed, actual glance direction of truck drivers with annotated direction. Findings indicate that while for some locations high annotation accuracy is achieved, for most locations accuracy is well below 50%. Higher accuracy can be obtained by clustering these locations, but this also leads to reduced detail of the annotation, suggesting that decisions to use clustering should take into account the purpose of the annotation. The data also show that high agreement between annotators does not guarantee high accuracy. We argue that the accuracy of annotation needs to be verified experimentally more often.

Dataset Construction from Naturalistic Driving in Roundabouts.

A proper driver characterization in complex environments using computational techniques depends on the richness and variety of data obtained from naturalistic driving. The present article proposes the construction of a dataset from naturalistic driving specific to maneuvers in roundabouts and makes it open and available to the scientific community for performing their own studies. The dataset is a combination of data gathered from on-board instrumentation and data obtained from the post-processing of maps as well as recorded videos. The approach proposed in this paper consists of handling roundabouts as a stretch of road that includes 100 m before the entrance, the internal part, and 100 m after the exit. This stretch of road is then spatially sampled in small sections to which data are associated.

Think Aloud Protocol Applied in Naturalistic Driving for Driving Rules Generation.

Understanding naturalistic driving in complex scenarios is an important step towards autonomous driving, and several approaches have been adopted for modeling driver's behaviors. This paper presents the methodology known as "Think Aloud Protocol" to model driving. This methodology is a data-gathering technique in which drivers are asked to verbalize their thoughts as they are driving which are then recorded, and the ensuing analysis of the audios and videos permits to derive driving rules. The goal of this paper is to show how think aloud methodology is applied in the naturalistic driving area, and to demonstrate the validity of the proposed approach to derive driving rules. The paper presents, firstly, the background of the think aloud methodology and then presents the application of this methodology to driving in roundabouts. The general deployment of this methodology consists of several stages: driver preparation, data collection, audio and video processing, generation of coded transcript files, and the generation of driving rules. The main finding of this study is that think aloud protocol can be applied to naturalistic driving, and even some potential limitations as discussed in the paper, the presented methodology is a relatively easy approach to derive driving rules.

Research on the Influence of Vehicle Speed on Safety Warning Algorithm: A Lane Change Warning System Case Study.

Speed has an important impact on driving safety, however, this factor is not included in existing safety warning algorithms. This study uses lane change systems to study the influence of vehicle speed on safety warning algorithms, aiming to determine lane change warning rules for different speeds (DS-LCW). Thirty-five drivers are recruited to carry out an extreme trial and naturalistic driving experiment. The vehicle speed, relative speed, relative distance, and minimum safety deceleration (MSD) related to lane change characteristics are then analyzed and calculated as warning rule characterization parameters. Lane change warning rules for a rear vehicle in the target lane under four-speed levels of 60 ≤ v < 70 km/h, 70 ≤ v < 80 km/h, 80 ≤ v < 90 km/h, and v ≥ 90 km/h are established. The accuracy of lane change warning rules not considering speed level (NDS-LCW) and ISO 17387 are found to be 87.5% and 79.8%, respectively. Comparatively, the accuracy rate of DS-LCW under four-speed levels is 94.6%, 93.8%, 90.0%, and 92.6%, respectively, which is significantly superior. The algorithm proposed in this paper provides warning in the lane change process with a smaller relative distance, and the accuracy rate of DS-LCW is significantly superior to NDS-LCW and ISO 17387.

Practical and statistical challenges in driving research.

Driving is an integral aspect of many modern societies, and motor vehicle safety is an important public health issue. With advances in sensor technology, more and more driving data are being collected by researchers, insurers, and automobile companies, which has increased the need and opportunities for statisticians to be involved in driving research. This report discusses several practical and statistical challenges in driver-level studies, including the process of defining meaningful driving metrics, issues related to "Big Data" aspects of driving research, and the principle of reproducible research.

Semiparametric Bayesian models for evaluating time-variant driving risk factors using naturalistic driving data and case-crossover approach.

Driver behavior is a major contributing factor for traffic crashes, a leading cause of death and injury in the United States. The naturalistic driving study (NDS) revolutionizes driver behavior research by using sophisticated nonintrusive in-vehicle instrumentation to continuously record driving data. This paper uses a case-crossover approach to evaluate driver-behavior risk. To properly model the unbalanced and clustered binary outcomes, we propose a semiparametric hierarchical mixed-effect model to accommodate both among-strata and within-stratum variations. This approach overcomes several major limitations of the standard models, eg, constant stratum effect assumption for conditional logistic model. We develop 2 methods to calculate the marginal conditional probability. We show the consistency of parameter estimation and asymptotic equivalence of alternative estimation methods. A simulation study indicates that the proposed model is more efficient and robust than alternatives. We applied the model to the 100-Car NDS data, a large-scale NDS with 102 participants and 12-month data collection. The results indicate that cell phone dialing increased the crash/near-crash risk by 2.37 times (odds ratio: 2.37, 95% CI, 1.30-4.30) and drowsiness increased the risk 33.56 times (odds ratio: 33.56, 95% CI, 21.82-52.19). This paper provides new insight into driver behavior risk and novel analysis strategies for NDS studies.

Objective Monitoring of Driving Behavior in Parkinson's Disease: The Utility of the Chula Parkinson Car®.

INTRODUCTION: Driving competency is important to evaluate among individuals with Parkinson's disease (PD). Driving in natural situations is the preferred assessment method; thus, we used a naturalistic driving environment to identify driving competency among individuals with PD in comparison to healthy age-matched controls. METHODS: Based on a power analysis, we recruited 20 participants (10 with PD and 10 healthy age-matched controls). Each participant completed 3 tasks while driving the ChulaPD car, a 4-door sedan installed with computerized monitoring systems. The tasks were forward and backward vehicle movement, reversing into a parking space, and parking parallel to a sidewalk. Trip start and end times, vehicle speed, and acceleration and deceleration times were logged using steering wheel motion, location parking sensors, and dashboard cameras and compared between groups. RESULTS: Age, gender, possession of a driver's license, present driving conditions, Thai Mini-Mental State Examination score, and driving experience did not significantly differ between groups. However, the PD group took longer to complete the driving tests (p = 0.002), had slower vehicle speeds (p = 0.002), longer brake times (p = 0.007), and decreased brake pressure ability (p = 0.009). Under normalized conditions, the ratio of failed driver's license tests was also higher among the PD group than in the control group (70 vs. 10%, p = 0.006). CONCLUSIONS: Individuals with PD had less-than-adequate driving ability based on our naturalistic setting. Our assessment method may be useful in other populations with chronic illnesses or for older adults. We discuss how naturalistic assessments could become the standard for evaluating driving ability in Thailand and elsewhere.

Driver crash risk factors and prevalence evaluation using naturalistic driving data.

The accurate evaluation of crash causal factors can provide fundamental information for effective transportation policy, vehicle design, and driver education. Naturalistic driving (ND) data collected with multiple onboard video cameras and sensors provide a unique opportunity to evaluate risk factors during the seconds leading up to a crash. This paper uses a National Academy of Sciences-sponsored ND dataset comprising 905 injurious and property damage crash events, the magnitude of which allows the first direct analysis (to our knowledge) of causal factors using crashes only. The results show that crash causation has shifted dramatically in recent years, with driver-related factors (i.e., error, impairment, fatigue, and distraction) present in almost 90% of crashes. The results also definitively show that distraction is detrimental to driver safety, with handheld electronic devices having high use rates and risk.

The Alabama VIP older driver study rationale and design: examining the relationship between vision impairment and driving using naturalistic driving techniques.

BACKGROUND: Older drivers aged ≥70 years old have among the highest rates of motor vehicle collisions (MVC) compared to other age groups. Driving is a highly visual task, and older adults have a high prevalence of vision impairment compared to other ages. Most studies addressing visual risk factors for MVCs by older drivers utilize vehicle accident reports as the primary outcome, an approach with several methodological limitations. Naturalistic driving research methods overcome these challenges and involve installing a high-tech, unobtrusive data acquisition system (DAS) in an older driver's own vehicle. The DAS continuously records multi-channel video of driver and roadway, sensor-based kinematics, GPS location, and presence of nearby objects in front of the vehicle, providing an objective measure of driving exposure. In this naturalistic driving study, the purpose is to examine the relationship between vision and crashes and near-crashes, lane-keeping, turning at intersections, driving performance during secondary tasks demands, and the role of front-seat passengers. An additional aim is to compare results of the on-road driving evaluation by a certified driving rehabilitation specialist to objective indicators of driving performance derived from the naturalistic data. METHODS: Drivers ≥70 years old are recruited from ophthalmology clinics and a previous population-based study of older drivers, with the goal of recruiting persons with wide ranging visual function. Target samples size is 195 drivers. At a baseline visit, the DAS is installed in the participant's vehicle and a battery of health and functional assessments are administered to the driver including visual-sensory and visual-cognitive tests. The DAS remains installed in the vehicle for six months while the participant goes about his/her normal driving with no imposed study restrictions. After six months, the driver returns for DAS de-installation, repeat vision testing, and an on-road driving evaluation by a certified driving rehabilitation specialist (CDRS). The data streams recorded by the DAS are uploaded to the data coordinating center for analysis. DISCUSSION: The Alabama VIP Older Driver Study is the first naturalistic older driver study specifically focused on the enrollment of drivers with vision impairment in order to study the relationship between visual dysfunction and driver safety and performance.

Using Topic Modeling to Develop Multi-level Descriptions of Naturalistic Driving Data from Drivers with and without Sleep Apnea.

One challenge in using naturalistic driving data is producing a holistic analysis of these highly variable datasets. Typical analyses focus on isolated events, such as large g-force accelerations indicating a possible near-crash. Examining isolated events is ill-suited for identifying patterns in continuous activities such as maintaining vehicle control. We present an alternative approach that converts driving data into a text representation and uses topic modeling to identify patterns across the dataset. This approach enables the discovery of non-linear patterns, reduces the dimensionality of the data, and captures subtle variations in driver behavior. In this study topic models are used to concisely described patterns in trips from drivers with and without untreated obstructive sleep apnea (OSA). The analysis included 5000 trips (50 trips from 100 drivers; 66 drivers with OSA; 34 comparison drivers). Trips were treated as documents, and speed and acceleration data from the trips were converted to "driving words." The identified patterns, called topics, were determined based on regularities in the co-occurrence of the driving words within the trips. This representation was used in random forest models to predict the driver condition (i.e., OSA or comparison) for each trip. Models with 10, 15 and 20 topics had better accuracy in predicting the driver condition, with a maximum AUC of 0.73 for a model with 20 topics. Trips from drivers with OSA were more likely to be defined by topics for smaller lateral accelerations at low speeds. The results demonstrate topic modeling as a useful tool for extracting meaningful information from naturalistic driving datasets.

Detection of Lane-Change Events in Naturalistic Driving Videos.

Lane changes are important behaviors to study in driving research. Automated detection of lane-change events is required to address the need for data reduction of a vast amount of naturalistic driving videos. This paper presents a method to deal with weak lane-marker patterns as small as a couple of pixels wide. The proposed method is novel in its approach to detecting lane-change events by accumulating lane-marker candidates over time. Since the proposed method tracks lane markers in temporal domain, it is robust to low resolution and many different kinds of interferences. The proposed technique was tested using 490 h of naturalistic driving videos collected from 63 drivers. The lane-change events in a 10-h video set were first manually coded and compared with the outcome of the automated method. The method's sensitivity was 94.8% and the data reduction rate was 93.6%. The automated procedure was further evaluated using the remaining 480-h driving videos. The data reduction rate was 97.4%. All 4971 detected events were manually reviewed and classified as either true or false lane-change events. Bootstrapping showed that the false discovery rate from the larger data set was not significantly different from that of the 10-h manually coded data set. This study demonstrated that the temporal processing of lane markers is an effcient strategy for detecting lane-change events involving weak lane-marker patterns in naturalistic driving.