AI-Based Vehicular Network toward 6G and IoT: Deep Learning Approaches

In recent years, vehicular networks have become increasingly large, heterogeneous, and dynamic, making it difficult to meet strict requirements of ultralow latency, high reliability, high security, and massive connections for next generation (6G) networks. Recently, deep learning (DL ) has emerged as a powerful artificial intelligence (AI ) technique to optimize the efficiency and adaptability of vehicle and wireless communication. However, rapidly increasing absolute numbers of vehicles on the roads are leading to increased automobile accidents, many of which are attributable to drivers interacting with their mobile phones. To address potentially dangerous driver behavior, this study applies deep learning approaches to image recognition to develop an AI-based detection system that can detect potentially dangerous driving behavior. Multiple convolutional neural network (CNN )-based techniques including VGG16, VGG19, Densenet, and Openpose were compared in terms of their ability to detect and identify problematic driving.

The Effects on Driving Behavior When Using a Head-mounted Display in a Dynamic Driving Simulator

Driving simulators are established tools used during automotive development and research. Most simulators use either monitors or projectors as their primary display system. However, the emergence of a new generation of head-mounted displays has triggered interest in using these as the primary display type. The general benefits and drawbacks of head-mounted displays are well researched, but their effect on driving behavior in a simulator has not been sufficiently quantified. This article presents a study of driving behavior differences between projector-based graphics and head-mounted display in a large dynamic driving simulator. This study has selected five specific driving maneuvers suspected of affecting driving behavior differently depending on the choice of display technology. Some of these maneuvers were chosen to reveal changes in lateral and longitudinal driving behavior. Others were picked for their ability to highlight the benefits and drawbacks of head-mounted displays in a driving context. The results show minor changes in lateral and longitudinal driver behavior changes when comparing projectors and a head-mounted display. The most noticeable difference in favor of projectors was seen when the display resolution is critical to the driving task. The choice of display type did not affect simulator sickness nor the realism rated by the subjects.

A Recognition Method of Aggressive Driving Behavior Based on Ensemble Learning

Aggressive driving behavior (ADB) is one of the main causes of traffic accidents. The accurate recognition of ADB is the premise to timely and effectively conduct warning or intervention to the driver. There are some disadvantages, such as high miss rate and low accuracy, in the previous data-driven recognition methods of ADB, which are caused by the problems such as the improper processing of the dataset with imbalanced class distribution and one single classifier utilized. Aiming to deal with these disadvantages, an ensemble learning-based recognition method of ADB is proposed in this paper. First, the majority class in the dataset is grouped employing the self-organizing map (SOM) and then are combined with the minority class to construct multiple class balance datasets. Second, three deep learning methods, including convolutional neural networks (CNN), long short-term memory (LSTM), and gated recurrent unit (GRU), are employed to build the base classifiers for the class balance datasets. Finally, the ensemble classifiers are combined by the base classifiers according to 10 different rules, and then trained and verified using a multi-source naturalistic driving dataset acquired by the integrated experiment vehicle. The results suggest that in terms of the recognition of ADB, the ensemble learning method proposed in this research achieves better performance in accuracy, recall, and F1-score than the aforementioned typical deep learning methods. Among the ensemble classifiers, the one based on the LSTM and the Product Rule has the optimal performance, and the other one based on the LSTM and the Sum Rule has the suboptimal performance.

A Collision Relationship-Based Driving Behavior Decision-Making Method for an Intelligent Land Vehicle at a Disorderly Intersection via DRQN

An intelligent land vehicle utilizes onboard sensors to acquire observed states at a disorderly intersection. However, partial observation of the environment occurs due to sensor noise. This causes decision failure easily. A collision relationship-based driving behavior decision-making method via deep recurrent Q network (CR-DRQN) is proposed for intelligent land vehicles. First, the collision relationship between the intelligent land vehicle and surrounding vehicles is designed as the input. The collision relationship is extracted from the observed states with the sensor noise. This avoids a CR-DRQN dimension explosion and speeds up the network training. Then, DRQN is utilized to attenuate the impact of the input noise and achieve driving behavior decision-making. Finally, some comparative experiments are conducted to verify the effectiveness of the proposed method. CR-DRQN maintains a high decision success rate at a disorderly intersection with partially observable states. In addition, the proposed method is outstanding in the aspects of safety, the ability of collision risk prediction, and comfort.

Influence of Landscape Design on Driving Behavior Based on Slope Illusion

In a continuous downhill section of a mountain highway, factors such as road alignment, roadside environment, and other visual characteristics will impact the slope illusion drivers experience and engage in unsafe driving behaviors. To improve the negative consequences of slope illusion and driving safety in continuous downhill sections, the effects of plant spacing, height, roadside distance, and color on driving behavior were all studied by simulating the plant landscape in a virtual environment. A driving simulator and UC-win/road software were used to conduct an indoor driving simulation experiment, and parameters such as speed and lateral position offset were used as the evaluation indices of driving stability to reflect the driver’s speed perception ability with subjective equivalent speeds. The results show that a plant landscape with appropriate plant spacing, height, roadside separation, and color is conducive to improving driving stability. Furthermore, a landscape with a height of 3 m, spacing of 10 m, roadside spacing of 0.75 m, and appropriate color matching can enhance the slope perception ability and speed perception ability of drivers, which is conducive to improving the driving safety of continuous downhill sections.

Understanding Traffic Accidents among Young Drivers in Qatar

Young drivers are generally associated with risky driving behaviors that can lead to crash involvement. Many self-report measurement scales are used to assess such risky behaviors. This study is aimed to understand the risky driving behaviors of young adults in Qatar and how such behaviors are associated with crash involvement. This was achieved through the usage of validated self-report measurement scales adopted for the Arabic context. A nationwide cross-sectional and exploratory study was conducted in Qatar from January to April 2021. Due to the Covid-19 pandemic, the survey was conducted online. Therefore, respondents were selected conveniently. Hence, the study adopted a non-probability sampling method in which convenience and snowball sampling were used. A total of 253 completed questionnaires were received, of which 57.3% were female, and 42.7% were male. Approximately 55.8% of these young drivers were involved in traffic accidents after obtaining their driving license. On average, most young drivers do have some risky driving behavior accompanied by a low tendency to violate traffic laws, and their driving style is not significantly controlled by their personality on the road. The older young drivers are more involved in traffic accidents than the younger drivers, i.e., around 1.5 times more likely. Moreover, a young male driver is 3.2 times less likely to be involved in traffic accidents than a female driver. In addition, males are only 0.309 times as likely as females to be involved in an accident and have approximately a 70% lower likelihood of having an accident versus females. The analysis is complemented with the association between young drivers’ demographic background and psychosocial-behavioral parameters (linking risky driving behavior, personality, and obligation effects on crash involvement). Some interventions are required to improve driving behavior, such as driving apps that are able to monitor and provide corrective feedback.