Search Results

The optimization and further development of automated driving functions offers great potential to relieve the driver in various driving situations and increase road safety. Simulative testing in particular is an indispensable tool in this process, allowing conclusions to be drawn about the design of automated driving functions at a very early stage of development. In this context, the use of driving simulators provides support so that the driving functions of tomorrow can be experienced in a very safe and reproducible environment. The focus of the acceptance and optimization of automated driving functions is particularly on vehicle lateral control functions. As part of this paper, a test person study was carried out regarding manual vehicle lateral control on the dynamic vehicle road simulator at the Institute of Automotive Engineering.

Investigating human driver behavior enhances the acceptance of the autonomous driving and increases road safety in heterogeneous environments with human-operated and autonomous vehicles. The previously established driver fingerprint model, focuses on the classification of driving style based on CAN bus signals. However, driving styles are inherently complex and influenced by multiple factors, including changing driving environments and driver states. To comprehensively create a driver profile, an in-car measurement system based on the Driver-Driven vehicle-Driving environment (3D) framework is developed. The measurement system records emotional and physiological signals from the driver, including ECG signal and heart rate. A Raspberry Pi camera is utilized on the dashboard to capture the driver's facial expressions and a trained convolutional neural network (CNN) recognizes emotion. To conduct unobtrusive ECG measurements, an ECG sensor is integrated into the steering wheel.

As part of the safety validation of advanced driver assistance systems (ADAS) and automated driving (AD) functions, it is necessary to demonstrate that the frequency at which the system exhibits hazardous behavior (HB) in the field is below an acceptable threshold. This is typically tested by observation of the system behavior in a field operational test (FOT). For situations in which the system under test (SUT) actively intervenes in the dynamic driving behavior of the vehicle, it is assessed whether the SUT exhibits HB. Since the accepted threshold values are generally small, the amount of data required for this strategy is usually very large. This publication proposes an approach to reduce the amount of data required for the evaluation of emergency intervention systems with a state machine based intervention logic by including the time periods between intervention events in the validation process.

Understanding left-turn vehicle-pedestrian accident mechanisms is critical for developing accident-prevention systems. This study aims to clarify the features of driver behavior focusing on drivers’ gaze, vehicle speed, and time to collision (TTC) during left turns at intersections on left-hand traffic roads. Herein, experiments with a sedan and light-duty truck (< 7.5 tons GVW) are conducted under four conditions: no pedestrian dummy (No-P), near-side pedestrian dummy (Near-P), far-side pedestrian dummy (Far-P) and near-and-far side pedestrian dummies (NF-P). For NF-P, sedans have a significantly shorter gaze time for left-side mirrors compared with light-duty trucks. The light-duty truck’s average speed at the initial line to the intersection (L1) and pedestrian crossing line (L0) is significantly lower than the sedan’s under No-P, Near-P, and NF-P conditions, without any significant difference between any two conditions.

It is an important factor in electric vehicles to show customers how much they can drive with the energy of the remaining battery. If the remaining mileage is not accurate, electric vehicle drivers will have no choice but have to feel anxious about the mileage. Additionally, the potential customers have range anxiety when they consider Electric Vehicles. If the remaining mileage to drive is wrong, drivers may not be able to get to the charging station and may not be able to drive because the battery runs out. It is important to show the remaining available driving range exactly for drivers. The previous study proposed an advanced model by predicting the remaining mileage based on actual driving data and based on reflecting the pattern of customers who drive regularly. The Bayesian linear regression model was right model in previous study.

Accurately predicting real-world vehicle energy consumption is essential for optimizing vehicle designs, enhancing energy efficiency, and developing effective energy management strategies. This paper presents a data-driven approach that utilizes machine learning techniques and a comprehensive dataset of vehicle parameters and environmental factors to create precise energy consumption prediction models. The methodology involves recording real-world vehicle data using data loggers to extract information from the CAN bus systems for ICE and hybrid electric, as well as hydrogen and battery fuel cell vehicles. Data cleaning and cycle-based analysis are employed to process the dataset for accurate energy consumption prediction. This includes cycle detection and analysis using methods from statistics and signal processing, and then pattern recognition based on these metrics.

Highway safety remains a significant concern, especially in mixed traffic scenarios involving heavy-duty vehicles (HDV) and smaller passenger cars. The vulnerability of HDVs following closely behind smaller cars is evident in incidents involving the lead vehicle, potentially leading to catastrophic rear-end collisions. This paper explores how automatic speed enforcement systems, using speed cameras, can mitigate risks for HDVs in such critical situations. While historical crash data consistently demonstrates the reduction of accidents near speed cameras, this paper goes beyond the conventional notion of crash occurrence reduction. Instead, it investigates the profound impact of driver behavior changes within desired travel speed distribution, especially around speed cameras, and their contribution to the safety of trailing vehicles, with a specific focus on heavy-duty trucks in accident-prone scenarios.

A fully instrumented Tesla Model 3 was used to collect thousands of hours of real-world automated driving data, encompassing both Autopilot and Full Self-Driving modes. This comprehensive dataset included vehicle operational parameters from the data busses, capturing details such as powertrain performance, energy consumption, and the control of advanced driver assistance systems (ADAS). Additionally, interactions with the surrounding traffic were recorded using a perception kit developed in-house equipped with LIDAR and a 360-degree camera system. We collected the data as part of a larger program to assess energy-efficient driving behavior of production connected and automated vehicles. One important aspect of characterizing the test vehicle is predicting its car-following behavior. Using both uncontrolled on-road tests and dedicated tests with a lead car performing set speed maneuvers, we tuned conventional adaptive cruise control (ACC) equations to fit the vehicle’s behavior.

Human driving behavior's inherent variability, randomness, individual differences, and dynamic vehicle-road situations give human-machine cooperative (HMC) driving considerable uncertainty, which affects the applicability and effectiveness of HMC control in complex scenes. To overcome this challenge, we present a novel data-enabled game output regulation approach for HMC driving. Firstly, a global human-vehicle-road (HVR) model is established considering the varied driver's steering characteristic parameters, such as delay time, preview time, and steering gain, as well as the uncertainty of tire cornering stiffness and variable road curvature disturbance. The robust output regulation theory has been employed to ensure the global DVR system's closed-loop stability, asymptotic tracking, and disturbance rejection, even with an unknown driver's internal state. Secondly, an interactive shared steering controller has been designed to provide personalized driving assistance.

Driver's driving style has a great impact on lane changing behavior, especially in scenarios such as freeway on-ramps that contain a strong willingness to change lanes, both in terms of inter-vehicle interactions during lane changing and in terms of the driving styles of the two vehicles. This paper proposes a study on game-theoretic decision-making for lane-changing on highway on-ramps considering driving styles, aiming to facilitate safer and more efficient merging while adequately accounting for driving styles. Firstly, the six features proposed by the EXID dataset of lane-changing vehicles were subjected to Principal Component Analysis (PCA) and the three principal components after dimensionality reduction were extracted, and then clustered according to the principal components by the K-means algorithm. The parameters of lane-changing game payoffs are computed based on the clustering centers under several styles.

The steer-by-wire (SBW) system, an integral component of the drive-by-wire chassis responsible for controlling the lateral motion of a vehicle, plays a pivotal role in enhancing vehicle safety. However, it poses a unique challenge concerning steering wheel return control, primarily due to its fundamental characteristic of severing the mechanical connection between the steering wheel and the turning wheel. This disconnect results in the inability to directly transmit the self-aligning torque to the steering wheel, giving rise to complications in ensuring a seamless return process. In order to realize precise control of steering wheel return, solving the problem of insufficient low-speed return and high-speed return overshoot of the steering wheel of the SBW system, this paper proposes a steering wheel active return control strategy for SBW system based on the backstepping control method.

Data-driven driving safety assessment is crucial in understanding the insights of traffic accidents caused by dangerous driving behaviors. Meanwhile, quantifying driving safety through well-defined metrics in real-world naturalistic driving data is also an important step for the operational safety assessment of automated vehicles (AV). However, the lack of flexible data acquisition methods and fine-grained datasets has hindered progress in this critical area. In response to this challenge, we propose a novel dataset for driving safety metrics analysis specifically tailored to car-following situations. Leveraging state-of-the-art Artificial Intelligence (AI) technology, we employ drones to capture high-resolution video data at 12 traffic scenes in the Phoenix metropolitan area. After that, we developed advanced computer vision algorithms and semantically annotated maps to extract precise vehicle trajectories and leader-follower relations among vehicles.

Compared with urban areas, the road surface in mountainous areas generally has a larger slope, larger curvature and narrower width, and the vehicle may roll over and other dangers on such a road. In the case of limited driver information, if the two cars on the mountain road approach fast, it is very likely to occur road blockage or even collision. Multi-vehicle cooperative control technology can integrate the driving data of nearby vehicles, expand the perception range of vehicles, assist driving through multi-objective optimization algorithm, and improve the driving safety and traffic system reliability. Most existing studies on cooperative control of multiple vehicles is mainly focused on urban areas with stable environment, while ignoring complex conditions in mountainous areas and the influence of driver status. In this study, a digital twin based multi-vehicle cooperative warning system was proposed to improve the safety of multiple vehicles on mountain roads.

As a key tool to maintain urban cleanliness and improve the road environment, road cleaning vehicles play an important role in improving the quality of life of residents. However, the traditional road cleaning vehicle requires the driver to monitor the situation of road garbage at all times and manually operate the cleaning process, resulting in an increase in the driver 's work intensity. To solve this problem, this paper proposes a road garbage recognition algorithm based on improved YOLOv5, which aims to reduce labor consumption and improve the efficiency of road cleaning. Firstly, the lightweight network MobileNet-V3 is used to replace the backbone feature extraction network of the YOLOv5 model. The number of parameters and computational complexity of the model are greatly reduced by replacing the standard convolution with the deep separable convolution, which enabled the model to have faster reasoning speed while maintaining higher accuracy.

Battery Run-down under the Electric Vehicle Operation (BREVO) model is a model that links the driver’s travel pattern to physics-based battery degradation and powertrain energy consumption models. The model simulates the impacts of charging behavior, charging rate, driving patterns, and multiple energy management modules on battery capacity degradation. This study implements reinforcement learning (RL) to the simplified BREVO model to optimize drivers’ decisions on charging such as charging rate, charging time, and charging capacity needed. This is done by a reward function that considers both the driver’s daily travel demands and the minimization of battery degradation over a year. It shows that using appropriate charger type (No Charge, Level 1, Level 2, direct-current Fast Charge [DCFC], extreme Fast Charging [xFC]) with an appropriate charging time can reduce battery degradation and total charging cost at the end of the year while satisfying driver’s daily travel demand.

Driver state monitoring is a crucial technology for enhancing road safety and preventing human error-caused accidents in the era of autonomous vehicles. This paper presents CogniSafe, a comprehensive driver monitoring system that uses deep learning and computer vision methods to detect various types of driver distractions and fatigue. CogniSafe consists of four modules: Driver anomaly detection and classification: A novel two-phase network that proposes and recognizes driver anomalies, such as texting, drinking, and adjusting radios, using multimodal and multiview input. Gaze estimation: A video-based neural network that jointly learns head pose and gaze dynamics, achieving robust and efficient gaze estimation across different head poses. Eye state analysis: A multi-tasking CNN that encodes features from both eye and mouth regions, predicting the percentage of eye closure (PERCLOS) and the frequency of mouth opening (FOM).

With further development of autonomous vehicles additional challenges appear. One of these challenges arises in the context of mixed traffic scenarios where automated and autonomous vehicles coexist with manually operated vehicles as well as other road users such as cyclists and pedestrians. In this evolving landscape, understanding, predicting, and mimicking human driving behavior is becoming not only a challenging but also a compelling facet of autonomous driving research. This is necessary not only for safety reasons, but also to promote trust in artificial intelligence (AI), especially in self-driving cars where trust is often compromised by the opacity of neural network models. The central goal of this study is therefore to address this trust issue. A common approach to imitate human driving behavior through expert demonstrations is imitation learning (IL). However, balancing performance and explainability in these models is a major challenge.

Level 2 (L2) partial driving automation systems are rapidly emerging in the marketplace. L2 systems provide sustained automatic longitudinal and lateral vehicle motion control, reducing the need for drivers to continuously brake, accelerate and steer. Drivers, however, remain critically responsible for safely detecting and responding to objects and events. This paper summarizes variations of L2 systems (hands-on and/or hands-free) and considers human drivers’ roles when using L2 systems and for designing Human-Machine Interfaces (HMIs), including Driver Monitoring Systems (DMSs). In addition, approaches for examining potential unintended consequences of L2 usage and evaluating L2 HMIs, including field safety effect examination, are reviewed. The aim of this paper is to guide L2 system HMI development and L2 system evaluations, especially in the field, to support safe L2 deployment, promote L2 system improvements, and ensure well-informed L2 policy decision-making.

ADAS (Advanced Driver Assistance Systems) is a growing technology in automotive industry, intended to provide safety and comfort to the passengers with the help of variety of sensors like radar, camera, LIDAR etc. Though ADAS improved safety of passengers comparing to conventional non-ADAS vehicles, still it has some grey areas for safety enhancement and easy assistance to drivers. BSW (Blind Spot Warning) and LCA (Lane Change Assist) are ADAS function which assists the driver for lane changing. BSW alerts the driver about the vehicles which are in blind zone in adjacent lanes and LCA alerts the driver about approaching vehicles at a high velocity in adjacent lanes. In current ADAS systems, BSW and LCA alerts are given as optical and acoustic warnings which is placed in vehicle side mirrors. During lane change the driver must see the side mirrors to take a decision.

This paper presents a simulation approach to assess the impact of changes to the charge point infrastructure and policies on Electric Vehicle (EV) user satisfaction, combining both market drivers with the practicalities of EV usage. An agent-based model (ABM) approach is developed where a large number of EVs, that represent the user population, drive within a region of interest. By simulating the driver’s response to their charging experience, the model allows large scale trends to emerge from the population to guide infrastructure policies as the number of EVs increases beyond the initial early adopter market. The model incorporates a Monte Carlo approach to generate EV and driver agent instances with distinct characteristics, including battery size, vehicle type, driving style, sensitivity to range. The driver model is constructed to respond to events that may increase range anxiety, e.g. increasing the likelihood of charging as the driver becomes more anxious.