Drivers' Evaluation of Different Automated Driving Styles: Is It Both Comfortable and Natural?

OBJECTIVE

This study investigated users' subjective evaluation of three highly automated driving styles, in terms of comfort and naturalness, when negotiating a UK road in a high-fidelity, motion-based, driving simulator.

BACKGROUND

Comfort and naturalness play an important role in contributing to users' acceptance and trust of automated vehicles (AVs), although not much is understood about the types of driving style which are considered comfortable or natural.

METHOD

A driving simulator study, simulating roads with different road geometries and speed limits, was conducted. Twenty-four participants experienced three highly automated driving styles, two of which were recordings from human drivers, and the other was based on a machine learning (ML) algorithm, termed Defensive, Aggressive, and Turner, respectively. Participants evaluated comfort or naturalness of each driving style, for each road segment, and completed a Sensation Seeking questionnaire, which assessed their risk-taking propensity.

RESULTS

Participants regarded both human-like driving styles as more comfortable and natural, compared with the less human-like, ML-based, driving controller. Particularly, between the two human-like controllers, the Defensive style was considered more comfortable, especially for the more challenging road environments. Differences in preference for controller by driver trait were also observed, with the Aggressive driving style evaluated as more natural by the high sensation seekers.

CONCLUSION

Participants were able to distinguish between human- and machine-like AV controllers. A range of psychological concepts must be considered for the subjective evaluation of controllers.

APPLICATION

Insights into how different driver groups evaluate automated vehicle controllers are important in designing more acceptable systems.

Driver-Automated Vehicle Interaction in Mixed Traffic: Types of Interaction and Drivers' Driving Styles

OBJECTIVE

This study investigated drivers' subjective feelings and decision making in mixed traffic by quantifying driver's driving style and type of interaction.

BACKGROUND

Human-driven vehicles (HVs) will share the road with automated vehicles (AVs) in mixed traffic. Previous studies focused on simulating the impacts of AVs on traffic flow, investigating car-following situations, and using simulation analysis lacking experimental tests of human drivers.

METHOD

Thirty-six drivers were classified into three driver groups (aggressive, moderate, and defensive drivers) and experienced HV-AV interaction and HV-HV interaction in a supervised web-based experiment. Drivers' subjective feelings and decision making were collected via questionnaires.

RESULTS

Results revealed that aggressive and moderate drivers felt significantly more anxious, less comfortable, and were more likely to behave aggressively in HV-AV interaction than in HV-HV interaction. Aggressive drivers were also more likely to take advantage of AVs on the road. In contrast, no such differences were found for defensive drivers indicating they were not significantly influenced by the type of vehicles with which they were interacting.

CONCLUSION

Driving style and type of interaction significantly influenced drivers' subjective feelings and decision making in mixed traffic. This study brought insights into how human drivers perceive and interact with AVs and HVs on the road and how human drivers take advantage of AVs.

APPLICATION

This study provided a foundation for developing guidelines for mixed transportation systems to improve driver safety and user experience.

Driving behavior classification at signalized intersections using vehicle kinematics: Application of unsupervised machine learning

Driving behavior is considered as a unique driving habit of each driver and has a significant impact on road safety. This study proposed a novel data-driven Machine Learning framework that can classify driving behavior at signalized intersections considering two different signal conditions. To the best of our knowledge, this is the first study that investigates driving behavior at signalized intersections with two different conditions that are mostly used in practice, i.e., the control setting with the signal order of green-yellow-red and a flashing green setting with the signal order of green-flashing green-yellow-red. A driving simulator dataset collected from participants at Qatar University's Qatar Transportation and Traffic Safety Center, driving through multiple signalized intersections, was used. The proposed framework extracts volatility measures from vehicle kinematic parameters including longitudinal speed and acceleration. K-means clustering algorithm with elbow method was used as an unsupervised machine learning to cluster driving behavior into three classes (i.e., conservative, normal, and aggressive) and investigate the impact of signal conditions. The framework confirmed that in general driving behavior at a signalized intersection reflects drivers' habits and personality rather than the signal condition, still, it manifests the intersection nature that usually requires drivers to be more vigilant and cautious. Nonetheless, the results suggested that flashing green condition could make drivers more conservative, which could be due to the limited capabilities of human to estimate the remaining distance and the prolonged duration of the additional flashing green interval. The proposed framework and findings of the study were promising that can be used for clustering drivers into different styles for different conditions and might be beneficial for policymakers, researchers, and engineers.

Vehicle and Driver Monitoring System Using On-Board and Remote Sensors

This paper presents an integrated monitoring system for the driver and the vehicle in a single case of study easy to configure and replicate. On-board vehicle sensors and remote sensors are combined to model algorithms for estimating polluting emissions, fuel consumption, driving style and driver's health. The main contribution of this paper is the analysis of interactions among the above monitored features highlighting the influence of the driver in the vehicle performance and vice versa. This analysis was carried out experimentally using one vehicle with different drivers and routes and implemented on a mobile application. Compared to commercial driver and vehicle monitoring systems, this approach is not customized, uses classical sensor measurements, and is based on simple algorithms that have been already proven but not in an interactive environment with other algorithms. In the procedure design of this global vehicle and driver monitoring system, a principal component analysis was carried out to reduce the variables used in the training/testing algorithms with objective to decrease the transfer data via Bluetooth between the used devices: a biometric wristband, a smartphone and the vehicle's central computer. Experimental results show that the proposed vehicle and driver monitoring system predicts correctly the fuel consumption index in 84%, the polluting emissions 89%, and the driving style 89%. Indeed, interesting correlation results between the driver's heart condition and vehicular traffic have been found in this analysis.

Time-Series-Based Personalized Lane-Changing Decision-Making Model

In recent years, autonomous driving technology has been changing from "human adapting to vehicle" to "vehicle adapting to human". To improve the adaptability of autonomous driving systems to human drivers, a time-series-based personalized lane change decision (LCD) model is proposed. Firstly, according to the characteristics of the subject vehicle (SV) with respect to speed, acceleration and headway, an unsupervised clustering algorithm, namely, a Gaussian mixture model (GMM), is used to identify its three different driving styles. Secondly, considering the interaction between the SV and the surrounding vehicles, the lane change (LC) gain value is produced by developing a gain function to characterize their interaction. On the basis of the recognition of the driving style, this gain value and LC feature parameters are employed as model inputs to develop a personalized LCD model on the basis of a long short-term memory (LSTM) recurrent neural network model (RNN). The proposed method is tested using the US Open Driving Dataset NGSIM. The results show that the accuracy, F1 score, and macro-average area under the curve (macro-AUC) value of the proposed method for LC behavior prediction are 0.965, 0.951 and 0.983, respectively, and the performance is significantly better than that of other mainstream models. At the same time, the method is able to capture the LCD behavior of different human drivers, enabling personalized driving.

The Moderating Effects of Emotions on the Relationship Between Self-Reported Individual Traits and Actual Risky Driving Behaviors

Purpose

Researches addressing driving behaviors have not fully revealed how emotions affect risky driving behaviors and tend to focus on the effects of some negative emotions rather than those of more specific emotions. This study aimed to test the potential moderating effects of eight common driving emotions on the relationship between self-reported individual traits (sensation seeking and driving style) and actual risky driving behaviors, sequentially providing some implications for the risky driving behavior prevention.

Participants and Methods

A total of 78 licensed drivers were recruited from undergraduate students, company employees and taxi drivers in China. The participants’ data on self-reported driving style (SDBS) and self-reported sensation seeking (SSSS) were obtained through questionnaires. The participants’ data on actual risky driving behaviors (ARD) in eight driving emotional activation states were obtained through a series of emotion induction experiments and driving experiments. The Structural Equation Modeling (SEM) and moderating effect tests were employed to investigate the relationships of driving emotions, SDBS, SSSS and ARD.

Results

Results showed that anger and pleasure affected risky driving behaviors positively by enhancing the relationship between SDBS and ARD, while surprise and fear were negatively related to risky driving behaviors by weakening this relationship. Anxiety positively affected risky driving behaviors by synchronously enhancing the relationship between SDBS and ARD and the relationship between SSSS and ARD, while helplessness and relief affected risky driving behaviors negatively by weakening the two relationships. Contempt affected risky driving behaviors positively by enhancing the relation between SSSS and ARD.

Conclusion

The results illustrated the effects of different emotions on risky driving behaviors, and also partly explained the reasons for these effects. This research provided a source of reference for reducing traffic accidents caused by risky driving behaviors.

Assessing Drivers' Trust of Automated Vehicle Driving Styles With a Two-Part Mixed Model of Intervention Tendency and Magnitude

OBJECTIVE

This study examines how driving styles of fully automated vehicles affect drivers' trust using a statistical technique-the two-part mixed model-that considers the frequency and magnitude of drivers' interventions.

BACKGROUND

Adoption of fully automated vehicles depends on how people accept and trust them, and the vehicle's driving style might have an important influence.

METHOD

A driving simulator experiment exposed participants to a fully automated vehicle with three driving styles (aggressive, moderate, and conservative) across four intersection types (with and without a stop sign and with and without crossing path traffic). Drivers indicated their dissatisfaction with the automation by depressing the brake or accelerator pedals. A two-part mixed model examined how automation style, intersection type, and the distance between the automation's driving style and the person's driving style affected the frequency and magnitude of their pedal depression.

RESULTS

The conservative automated driving style increased the frequency and magnitude of accelerator pedal inputs; conversely, the aggressive style increased the frequency and magnitude of brake pedal inputs. The two-part mixed model showed a similar pattern for the factors influencing driver response, but the distance between driving styles affected how often the brake pedal was pressed, but it had little effect on how much it was pressed.

CONCLUSION

Eliciting brake and accelerator pedal responses provides a temporally precise indicator of drivers' trust of automated driving styles, and the two-part model considers both the discrete and continuous characteristics of this indicator.

APPLICATION

We offer a measure and method for assessing driving styles.

Holistic Vehicle Instrumentation for Assessing Driver Driving Styles

Nowadays, autonomous vehicles are increasing, and the driving scenario that includes both autonomous and human-driven vehicles is a fact. Knowing the driving styles of drivers in the process of automating vehicles is interest in order to make driving as natural as possible. To this end, this article presents a first approach to the design of a controller for the braking system capable of imitating the different manoeuvres that any driver performs while driving. With this aim, different experimental tests have been carried out with a vehicle instrumented with sensors capable of providing real-time information related to the braking system. The experimental tests consist of reproducing a series of braking manoeuvres at different speeds on a flat floor track following a straight path. The tests distinguish between three types of braking manoeuvre: maintained, progressive and emergency braking, which cover all the driving circumstances in which the braking system may intervene. This article presents an innovative approach to characterise braking types thanks to the methodology of analysing the data obtained by sensors during experimental tests. The characterisation of braking types makes it possible to dynamically classify three driving styles: cautious, normal and aggressive. The proposed classifications allow it possible to identify the driving styles on the basis of the pressure in the hydraulic brake circuit, the force exerted by the driver on the brake pedal, the longitudinal deceleration and the braking power, knowing in all cases the speed of the vehicle. The experiments are limited by the fact that there are no other vehicles, obstacles, etc. in the vehicle’s environment, but in this article the focus is exclusively on characterising a driver with methods that use the vehicle’s dynamic responses measured by on-board sensors. The results of this study can be used to define the driving style of an autonomous vehicle.

End-to-End Automated Lane-Change Maneuvering Considering Driving Style Using a Deep Deterministic Policy Gradient Algorithm

Changing lanes while driving requires coordinating the lateral and longitudinal controls of a vehicle, considering its running state and the surrounding environment. Although the existing rule-based automated lane-changing method is simple, it is unsuitable for unpredictable scenarios encountered in practice. Therefore, using a deep deterministic policy gradient (DDPG) algorithm, we propose an end-to-end method for automated lane changing based on lidar data. The distance state information of the lane boundary and the surrounding vehicles obtained by the agent in a simulation environment is denoted as the state space for an automated lane-change problem based on reinforcement learning. The steering wheel angle and longitudinal acceleration are used as the action space, and both the state and action spaces are continuous. In terms of the reward function, avoiding collision and setting different expected lane-changing distances that represent different driving styles are considered for security, and the angular velocity of the steering wheel and jerk are considered for comfort. The minimum speed limit for lane changing and the control of the agent for a quick lane change are considered for efficiency. For a one-way two-lane road, a visual simulation environment scene is constructed using Pyglet. By comparing the lane-changing process tracks of two driving styles in a simplified traffic flow scene, we study the influence of driving style on the lane-changing process and lane-changing time. Through the training and adjustment of the combined lateral and longitudinal control of autonomous vehicles with different driving styles in complex traffic scenes, the vehicles could complete a series of driving tasks while considering driving-style differences. The experimental results show that autonomous vehicles can reflect the differences in the driving styles at the time of lane change at the same speed. Under the combined lateral and longitudinal control, the autonomous vehicles exhibit good robustness to different speeds and traffic density in different road sections. Thus, autonomous vehicles trained using the proposed method can learn an automated lane-changing policy while considering safety, comfort, and efficiency.

Recognition of the Driving Style in Vehicle Drivers

This paper presents three different approaches to recognize driving style based on a hierarchical-model. Specifically, it proposes a hierarchical model for the recognition of the driving style for advanced driver-assistance systems (ADAS) for vehicles. This hierarchical model for the recognition of the style of the car driving considers three aspects: the driver emotions, the driver state, and finally, the driving style itself. In this way, the proposed hierarchical pattern is composed of three levels of descriptors/features, one to recognize the emotional states, another to recognize the driver state, and the last one to recognize the driving style. Each level has a set of descriptors, which can be sensed in a real context. Finally, the paper presents three driving style recognition algorithms based on different paradigms. One is based on fuzzy logic, another is based on chronicles (a temporal logic paradigm), and the last is based on an algorithm that uses the idea of the recognition process of the neocortex, called Ar2p (Algoritmo Recursivo de Reconocimiento de Patrones, for its acronym in Spanish). In the paper, these approaches are compared using real datasets, using different metrics of interest in the context of the Internet of the Things, in order to determine their capabilities of reasoning, adaptation, and the communication of information. In general, the initial results are encouraging, specifically in the cases of chronicles and Ar2p, which give the best results.

The Bulgarian Version of the Multidimensional Driving Style Inventory: Psychometric Properties

Road safety is one of the main priorities for the European Union. Different strategies and policies strive to increase the level of road safety across Europe and although this level has increased in the last couple of years the number of injuries and fatalities resulting from traffic accidents is still very high. The multidimensional driving style inventory (MDSI) is a self-reported instrument for the assessment of a person's habitual driving style and its connection to risky driving behaviour and involvement in different traffic accidents. The instrument was originally developed in Israel and there are several previous adaptations in different countries such as Argentina and Romania. The main objective of this study is to develop a valid and reliable version of the MDSI in Bulgaria. A study was conducted to evaluate the construction validity of the instrument and to test the validity of the factors in a Bulgarian sample (n = 456, male = 204; female = 252; average age = 37). Eight factors representing a specific driving style-dissociative, anxious, risky, angry, high-velocity, distress reduction, patient and careful and irrational-identified by previous versions of the MDSI are included in this study. The overall number of items in the inventory is 57 with Cronbach's alpha = 0.78. The current study in Bulgaria confirmed the structural organization of the initial version of the inventory. The results of the conducted study supported the reliability and validity of the Bulgarian version of the MDSI. The possible implementation of the instrument for the development of different programs for preventions and interventions is discussed here within.

Investigating the effect of personality on left-turn behaviors in various scenarios to understand the dynamics of driving styles

Objectives: To explore whether the driving style is static or dynamic, the purpose of this study was to investigate the effects of drivers' personality traits and driving scenarios on left-turn behaviors. We also evaluated drivers' perceptions in different situations to understand the occurrence of unreasonable driving behaviors.Methods: Two self-administered questionnaires were used to calculate participants' personality dimensions and driving style scores, respectively. Two experimental tasks were conducted by driving simulator with different weather, light, and speed of the opposing through vehicles.Results: Results showed that personality was significantly related to driving style. Neither the left-turn response process nor gap acceptance maneuvers were directly influenced by light conditions. In the test area, the average speed and the minimum speed of the left-turning vehicle, as well as the distance of the vehicle from the collision point when the speed is the smallest were significantly related to the weather, interaction type, and velocity difference, and the time to collision during which first brake or release gas pedal occurs was significantly related to velocity difference. Three out of the four driving conditions (except for light conditions) exhibited a statistically significant effect on the judgment of the arrival time of autonomous vehicles and their risk assessment, and personality was significantly related to risk assessment. Moreover, dangerous driving behavior was found to be significantly associated with drivers' driving personality.Conclusions: The driving style is dynamic and influenced by personality and driving conditions. Not only the wrong assessment of driving scenarios may lead to danger, but also the driver's personality characteristics such as self-interest, impulsively, and lack of self-confidence will lead to decision-making mistakes even when the judgment of driving scenarios is correct.

Data Anomaly Detection for Internet of Vehicles Based on Traffic Cellular Automata and Driving Style

The data validity of safe driving in the Internet of Vehicles (IoV) is the basis of improving the safety of vehicles. Different from a traditional information systems, the data anomaly analysis of vehicle safety driving faces the diversity of data anomaly and the randomness and subjectivity of the driver’s driving behavior. How to combine the characteristics of the IOV data with the driving style analysis to provide effective real-time anomaly data detection has become an important issue in the IOV applications. This paper aims at the critical safety data analysis, considering the large computing cost generated by the real-time anomaly detection of all data in the data package. We preprocess it through the traffic cellular automata model which is built to achieve the ideal abnormal detection effect with limited computing resources. On the basis of this model, the Anomaly Detection based on Driving style (ADD) algorithm is proposed to realize real-time and online detection of anomaly data related to safe driving. Firstly, this paper designs the driving coefficient and proposes a driving style quantization model to represent the driving style of the driver. Then, based on driving style quantization model and vehicle driving state information, a data anomaly detection algorithm is developed by using Gaussian mixture model (GMM). Finally, combining with the application scenarios of multi-vehicle collaboration in the Internet of Vehicles, this paper uses real data sets and simulation data sets to analyze the effectiveness of the proposed ADD algorithm.

An FPGA-Based Neuro-Fuzzy Sensor for Personalized Driving Assistance

Advanced driving-assistance systems (ADAS) are intended to automatize driver tasks, as well as improve driving and vehicle safety. This work proposes an intelligent neuro-fuzzy sensor for driving style (DS) recognition, suitable for ADAS enhancement. The development of the driving style intelligent sensor uses naturalistic driving data from the SHRP2 study, which includes data from a CAN bus, inertial measurement unit, and front radar. The system has been successfully implemented using a field-programmable gate array (FPGA) device of the Xilinx Zynq programmable system-on-chip (PSoC). It can mimic the typical timing parameters of a group of drivers as well as tune these typical parameters to model individual DSs. The neuro-fuzzy intelligent sensor provides high-speed real-time active ADAS implementation and is able to personalize its behavior into safe margins without driver intervention. In particular, the personalization procedure of the time headway (THW) parameter for an ACC in steady car following was developed, achieving a performance of 0.53 microseconds. This performance fulfilled the requirements of cutting-edge active ADAS specifications.

Driving Style Recognition Based on Electroencephalography Data From a Simulated Driving Experiment

Driving style is a very important indicator and a crucial measurement of a driver's performance and ability to drive in a safe and protective manner. A dangerous driving style would possibly result in dangerous behaviors. If the driving styles can be recognized by some appropriate classification methods, much attention could be paid to the drivers with dangerous driving styles. The driving style recognition module can be integrated into the advanced driving assistance system (ADAS), which integrates different modules to improve driving automation, safety and comfort, and then the driving safety could be enhanced by pre-warning the drivers or adjusting the vehicle's controlling parameters when the dangerous driving style is detected. In most previous studies, driver's questionnaire data and vehicle's objective driving data were utilized to recognize driving styles. And promising results were obtained. However, these methods were indirect or subjective in driving style evaluation. In this paper a method based on objective driving data and electroencephalography (EEG) data was presented to classify driving styles. A simulated driving system was constructed and the EEG data and the objective driving data were collected synchronously during the simulated driving. The driving style of each participant was classified by clustering the driving data via K-means. Then the EEG data was denoised and the amplitude and the Power Spectral Density (PSD) of four frequency bands were extracted as the EEG features by Fast Fourier transform and Welch. Finally, the EEG features, combined with the classification results of the driving data were used to train a Support Vector Machine (SVM) model and a leave-one-subject-out cross validation was utilized to evaluate the performance. The SVM classification accuracy was about 80.0%. Conservative drivers showed higher PSDs in the parietal and occipital areas in the alpha and beta bands, aggressive drivers showed higher PSD in the temporal area in the delta and theta bands. These results imply that different driving styles were related with different driving strategies and mental states and suggest the feasibility of driving style recognition from EEG patterns.

IoT On-Board System for Driving Style Assessment

The assessment of skills is essential and desirable in areas such as medicine, security, and other professions where mental, physical, and manual skills are crucial. However, often such assessments are performed by people called “experts” who may be subjective and are able to consider a limited number of factors and indicators. This article addresses the problem of the objective assessment of driving style independent of circumstances. The proposed objective assessment of driving style is based on eight indicators, which are associated with the vehicle’s speed, acceleration, jerk, engine rotational speed and driving time. These indicators are used to estimate three driving style criteria: safety, economy, and comfort. The presented solution is based on the embedded system designed according to the Internet of Things concept. The useful data are acquired from the car diagnostic port—OBD-II—and from an additional accelerometer sensor and GPS module. The proposed driving skills assessment method has been implemented and experimentally validated on a group of drivers. The obtained results prove the system’s ability to quantitatively distinguish different driving styles. The system was verified on long-route tests for analysis and could then improve the driver’s behavior behind the wheel. Moreover, the spider diagram approach that was used established a convenient visualization platform for multidimensional comparison of the result and comprehensive assessment in an intelligible manner.

Driving style indicator using UDRIVE NDS data

OBJECTIVE

In order to analyze specific events while driving (such as a safety critical event [SCE] or secondary task), we were interested in studying whether driving behavior was unusual around the event. An indicator characterizing driving style (driving style indicator [DSI]) was estimated for each driver by using naturalistic data. The analysis of the gap in driving style could be calculated for a specific trip or even a time window and could help characterize events: a more risky situation when DSI was above average, increase in safety margins when under average.

METHODS

Lateral acceleration and longitudinal acceleration were used for DSI calculation. The first step consisted in filtering the signal acquired by the inertial measurement unit (60 Hz). The noise was filtered out with an eighth order, phase-compensated digital low-pass Butterworth filter with a cut-out frequency of 5 Hz and offsets were compensated for. The second step consisted in calculating the jerk of the acceleration in lateral and longitudinal directions for each trip. The third step summarized the distribution of these jerks for all trips of each driver. Finally, the DSI was defined as the standard deviation of this distribution. A driver was represented by lateral DSI and longitudinal DSI.

RESULTS

The indicator was used on French pilot data (10 drivers) and on UK data (30 drivers) from the UDRIVE project. To assess this indicator, tests on track were conducted by professional drivers simulating two opposite driving style. The first results were promising and discriminated a smooth from a rough driving style. Indeed, in the pilot data, the classification was in accordance with our expectations and confirmed by videos. The same kind of distribution was observed in the UK data and needs to be confirmed when the UDRIVE database is complete.

CONCLUSION

DSI is a new parameter that will be used to define clusters of drivers and study variation of driving parameters in each class during selected events (SCE, secondary task, etc.) in the UDRIVE project.

A Review of Intelligent Driving Style Analysis Systems and Related Artificial Intelligence Algorithms

In this paper the various driving style analysis solutions are investigated. An in-depth investigation is performed to identify the relevant machine learning and artificial intelligence algorithms utilised in current driver behaviour and driving style analysis systems. This review therefore serves as a trove of information, and will inform the specialist and the student regarding the current state of the art in driver style analysis systems, the application of these systems and the underlying artificial intelligence algorithms applied to these applications. The aim of the investigation is to evaluate the possibilities for unique driver identification utilizing the approaches identified in other driver behaviour studies. It was found that Fuzzy Logic inference systems, Hidden Markov Models and Support Vector Machines consist of promising capabilities to address unique driver identification algorithms if model complexity can be reduced.