In-vehicle Systems Research Articles

Automatic Obstacle Avoidance Trolley System Based on Two-Dimensional ReS2/WSe2 Heterojunction Photodetector.

With the continuous advancement of intelligent vehicle technology, automatic obstacle avoidance (AOA) has become a crucial component of in-vehicle systems. This study introduces an AOA trolley system based on ReS2/WSe2 heterojunction photodetectors. The device with type-II band alignment exhibits an impressive on/off current ratio of 7.14 × 104, along with a rapid impulse response time of 32.3/24.4 μs, a detectivity of 2.0 × 1013 Jones, and an anisotropic ratio of 4.2 under 532 nm illumination. Additionally, the device is electrically tunable under an appropriate gate bias. These characteristics are ascribed to the engineering of the heterojunction interface between ReS2 and WSe2. Furthermore, this device was successfully integrated into an AOA system for autonomous vehicles. The system utilizes the high sensitivity and fast response features of two-dimensional materials, allowing it to effectively identify and avoid obstacles in real time, thereby enhancing driving safety.

Modeling and Control of Ejector-Based Hydrogen Circulation System for Proton Exchange Membrane Fuel Cell Systems

Ejector-based proton exchange membrane fuel cells (PEMFCs) are of great interest due to their simplicity and feasibility. Thus, proton exchange membrane fuel cells are considered the most suitable technology for in-vehicle systems, industrial applications, etc. Despite the passive characteristics of the ejector, active control of the hydrogen supply system is needed to ensure sufficient hydrogen, maintain the stack pressure, and ensure effective entrainment. In this research, a novel semi-empirical model is proposed to accurately predict the entrainment performance of the ejector with an 80 kW fuel cell system. According to the precise semi-empirical model, the hydrogen supply system and the anode channel are modeled. Then, a fuzzy logic controller (FLC) is developed to supply sufficient and adequate gas flow and maintain the rapid dynamic response. Compared to the conventional proportional–integral–derivative controller, the fuzzy logic controller could reduce the anode pressure variability by 5% during a stepped case and 2% during a dynamic case.

Artificial intelligence motivated flexible single-electrode mode multilayer triboelectric sensor for smart mobility systems

Triboelectric tactile sensing technique is increasingly coming into people’s lives to bring convenience, so that there is an urgent necessity for this technique to be applied to smart mobility to increase safety and enhance the mobility experience. Herein, a single-electrode mode multilayer triboelectric sensor (MTS) motivated by artificial intelligence (AI) is proposed, which consists of styrene-butadiene-styrene (SBS)/ poly (vinylida-ene fluoride-trifluoroethylene) (PVDF-TrFE) nanofibers (NFs) film as the friction layer and substrate, laser-induced graphene (LIG)/molybdenum disulfide (MoS2) as the charge trapping layer, and Ag as the electrode. The MTS exhibits remarkable sensing performance, such as a wide response range of 5–100 N, 0.4–2 Hz multi-frequency response capability, and non-contact sensing, as well as distinguished self-powered performance. To demonstrate the practical significance of the proposed MTS, two applications are explored specifically after equipping with AI, including: (i) a smart in-vehicle system is constructed, which consists of unlocking section and multifunctional control with early warning section. (ii) a smart car control system is implemented, which can carry out special tasks instead of humans. These applications provide reliable ways to promote the development of human-machine interaction and smart mobility, as well as help to make lifestyles and mobility smarter, safer and more convenient.

An automated learning method of semantic segmentation for train autonomous driving environment understanding

One of the major reasons for the explosion of autonomous driving in recent years is the great development of computer vision. As one of the most fundamental and challenging problems in autonomous driving, environment understanding has been widely studied. It directly determines whether the entire in-vehicle system can effectively identify surrounding objects of vehicles and make correct path planning. Semantic segmentation is the most important means of environment understanding among the many image recognition algorithms used in autonomous driving. However, the success of semantic segmentation models is highly dependent on human expertise in data preparation and hyperparameter optimization, and the tedious process of training is repeated over and over for each new scene. Automated machine learning (AutoML) is a research area for this problem that aims to automate the development of end-to-end ML models. In this paper, we propose an automatic learning method for semantic segmentation based on reinforcement learning (RL), which can realize automatic selection of training data and guide automatic training of semantic segmentation. The results show that our scheme converges faster and has higher accuracy than researchers manually training semantic segmentation models, while requiring no human involvement.

Multi-Layered Local Dynamic Map for a Connected and Automated In-Vehicle System

Automated Driving (AD) has been receiving considerable attention from industry, the public, and researchers for its ability to reduce accidents, emissions, and congestion. The purpose of this study is to extend the standardized Local Dynamic Map (LDM) by adding two new layers, and develop efficient and accurate algorithms designed to enhance AD by exploiting the LDM coupled with Cooperative Perception (CP). The LDM is implemented as a Neo4j graph database and extends the standard four-layer structure by adding a detection layer and a prediction layer. A custom Application Programming Interface (API) manages all incoming data, generates the LDM, and runs the algorithms. Currently, the API can match detected entities coming from different sources, correctly position them on the map even in the presence of high uncertainties in the data, and predict their future actions. We tested the developed LDM with real-world data, which we collected using a prototype vehicle from Centro Ricerche FIAT (CRF) Trento Branch—the supporting research center for this work—in urban, suburban, and highway areas of Trento, Italy. The results show that the developed solution is capable of accurately matching and predicting detected entities, is robust to high uncertainties in the data, and is efficient, achieving real-time performance in all scenarios. From these results we can conclude that the LDM and CP have the potential to be core parts of AD, bringing improvements to the development process.

Robustness of autonomous driving systems: extending the function failure modes taxonomy

While the development of autonomous driving systems (ADS) offers significant potential safety benefits, this is conditional upon addressing the challenges associated with increased complexity and multidisciplinarity, including the introduction of artificial intelligence (AI) embedded components. This increases the importance of assuring the robustness of these systems during their operations where possible hazards can arise from both the malfunctioning behavior of safety-related in-vehicle systems and functional insufficiencies or performance limitations of a system that is free from faults. Robustness is underpinned by the systematic consideration of failure modes which indicate how a system can potentially fail to deliver the intended function/performance and associated requirements. The introduction of AI components has made the identification for failures of ADSs challenging due to increased use of sensors and wireless communications which provide the backbone of core software functions of perception, planning and control. Current literature on failure modes mostly focuses on “physical” systems, and do not readily support the analysis of safety-critical systems which include AI components. This paper discusses the applicability of current failure mode taxonomies and their effectiveness in supporting the identification of failure modes within the context of design methodologies (such as FMEA) that underpin an effective approach to robustness of multidisciplinary autonomous driving systems.

Motion sickness countermeasures for autonomous driving: Trends and future directions

Motion Sickness (MS) remains a paramount concern in the evolving landscape of autonomous vehicles. As automation continues to reshape the passenger-vehicle interaction paradigm, the proliferation of diverse in-vehicle systems empowers passengers to disengage from the driving task. However, this new paradigm comes with the potential cost of exacerbating MS inside the car.In this work, we intended to: (1) present and review available countermeasures to prevent or mitigate MS, found in the literature, that could also be implemented inside a car; (2) identify current trends and gaps in countermeasures to MS; and (3) suggest future avenues of research for potential use-cases aiming to mitigate MS in autonomous driving.Through a comprehensive review of 65 publications, spanning from 2009 to 2023, we have organized existing literature into three distinctive categories and ten subcategories: (1) Vehicle-Centric Adaptation (Adaptation to Surface, Best Route ad Driving Style Adaptation), (2) In-Car Design and Environment (Design Guidelines for Seats, for Displays, for Windows and Climate Control), and (3) Sensory Cues (Visual, Haptic Audio and Olfactory Cues).Our findings suggest that Visual Cues and Motion Planning are the two strongest trends in MS countermeasures. In contrast, Olfactory Cues are the least studied approach. Our results also substantiate the viability of multimodal approaches as a promising solution for passengers in autonomous vehicles. The simultaneous application of various countermeasures might hold potential in mitigating MS effectively.As autonomous vehicles advance, these findings offer a strong basis for future research to decrease passenger's motion sickness and improve their well-being, safety, and comfort inside the car.

Galileo-Based Doppler Shifts and Time Difference Carrier Phase: A Static Case Demonstration.

The European Commission is designing and implementing new regulations for vehicle navigation in different sectors. Commission Delegated Regulation 2017/79 defines the compatibility and performance of the 112-based eCall in-vehicle systems. The regulation has a large impact on road transportation because it requires that all cars and light duty vehicles must be equipped with eCall devices. For heavy duty vehicles, a set of new regulations has been developed, starting from EU Regulation No 165/2014, in which the concept of smart tachographs was introduced to enforce the EU legislation on professional drivers' driving and resting times. In addition, intelligent speed assistance (ISA) devices increase the safety of road users. These new devices fully exploit the Global Navigation Satellite System (GNSS) to compute position velocity and time (PVT) information. In all these systems, the velocity of the vehicle plays a fundamental role; hence, a reliable and accurate velocity estimate is of utmost importance. In this work, two methods for velocity estimation using Galileo are presented and compared. The first exploits Doppler shift measurements, while the second uses time difference carrier phase (TDCP) measurements. The Doppler-based technique for velocity estimation is widely adopted in current devices, while the TDCP technique is emerging due to its promising high accuracy. The two methods are compared considering all the Galileo signals including E1, E5a, E5b, E5 Alt BOC and E6. The methods are compared in terms of velocity errors for both horizontal and vertical components using real static data. From the tests performed, it emerged that the TDCP has increased performance with respect to the Doppler-based solution. Among the Doppler-based solutions, the most accurate solution is the one obtained with the E5 Alt BOC signal.

A comparative analysis of voluntary in-vehicle distractions of drivers' visual attention

Distractions to the driving process, whether inside or outside of the vehicle, are becoming an increasing problem, especially distractions that are voluntary in nature which are avoidable but seem to be prevalent on our streets. A research gap was identified in the lack of comparability of in-vehicle distraction influences on drivers' visual attention. The paper presents research on the effects of in-vehicle distractions on drivers' visual attention, which was performed with 23 participants in a real environment (test track) during normal traffic conditions. The drivers were subjected to 15 in-vehicle distractions, including mobile phone use (handheld and hands-free), finding items inside the vehicle, and controlling and managing various in-vehicle systems. All distractions caused a significant reduction in visual attention to driving-related areas, including tasks such as setting the climate control or radio frequency, which are generally not perceived as disruptions. In fact, setting the climate control, navigational system, and radio frequency diverted visual attention for more than 40 % of the whole task duration. All tasks caused a reduction in the variability of gaze direction onto the windshield, indicating an occurrence of vision tunnelling, which was again very prominent even with tasks that are usually not perceived as very distracting, such as adjusting the side mirrors. The results, therefore, show that in-vehicle tasks that are voluntary in nature may significantly affect drivers' visual attention, and this is not limited to mobile phone and entertainment system use.

Analysis of the security of on-board information systems in vehicles

The features of the functioning of the on-board information systems of a car are considered. Threats to their security are analyzed, and methods for ensuring information security and functional security of on-board information systems are proposed. The design of road networks in the organization of road traffic is one of the factors in ensuring the functional security of modern intelligent transport systems, that is, compliance with such information security attributes as data confidentiality, integrity, availability, authenticity and novelty of data. The security of on-board vehicle information systems is a critical issue in the modern world, as more and more vehicles are equipped with electronic systems that may be vulnerable to cyber attacks. One of the main challenges of protecting on-board information systems is the wide range of devices and technologies used in modern vehicles. Different systems may have different security requirements and vulnerabilities. They may interact with each other in a complex way. Another challenge is that many of these systems were not designed originally with security in mind. They may lack basic security features such as encryption and authentication and use outdated software and protocols that are vulnerable to known attacks. The main types of attacks and threats to the elements of the transportation system that interact with the VANET were identified to analyze the information security of vehicle in-vehicle systems. Based on the theory of fuzzy sets under conditions of uncertainty and using the Fuzzy Logic Toolbox in the integrated Matlab environment, the level of information security of the OBU-VANET system was modeled. The obtained results allowed us to formulate the degree of information security of vehicle operation elements against unauthorized access to data. The results of the study showed that technical communication systems have the highest security level (> 0.7), and vehicles become the most vulnerable in public places.

An Extraction and Validity Evaluation Method Proposal for Monitoring Points for In-Vehicle Systems: Deriving Cybersecurity Requirements

To counter increasing threats of cyberattacks on automobiles, it is important to practice “secure by design.” For this reason, the regulations and standards that require a risk-based security design are being formulated in the automotive industry. In the concept phase described in ISO/SAE21434:2021, it is necessary to determine cybersecurity controls based on results of the threat analysis (TA) and the risk assessment (RA). On the other hand, the National Institute of Standards and Technology (NIST) Cybersecurity Framework (CSF) defines four categories of cybersecurity controls: protection, detection, response, and recovery demand to select cybersecurity controls from these categories. To be able to formulate appropriate cybersecurity control for detection, this article proposes an extension of the threat analysis and risk assessment (TARA) based on Japanese Automotive Standards Organization (JASO) TP15002:2016 to derive candidate monitoring points (MPs) to implement detection functions. We also present evaluation metrics for the MPs derived by our proposed method and discuss future research directions.

Divergent Selection Task Offloading Strategy for Connected Vehicles Based on Incentive Mechanism

With the improvements in the intelligent level of connected vehicles (CVs), travelers can enjoy services such as self-driving, self-parking and audiovisual entertainment inside the vehicle, which place extremely high demands on the computing power of onboard systems (OBSs). However, the arithmetic power of a single CV often cannot meet the diverse service demands of the in-vehicle system. As a new computing paradigm, task offloading based on vehicular edge computing has significant advantages in remedying the shortcomings of single-CV computing power and balancing the allocation of computing resources. This paper studied the computational task offloading of high-speed connected vehicles without the help of roadside edge servers in certain geographic areas. User vehicles (UVs) with insufficient computing power offload some of their computational tasks to nearby CVs with abundant resources. We explored the high-speed driving model and task classification model of CVs to refine the task offloading process. Additionally, inspired by game theory, we designed a divergent selection task offloading strategy based on an incentive mechanism (DSIM), in which we balanced the interests of both the user vehicle and service vehicles. CVs that contribute resources are rewarded to motivate more CVs to join. A DSIM algorithm based on a divergent greedy algorithm was introduced to maximize the total rewards of all volunteer vehicles while respecting the will of both the user vehicle and service vehicles. The experimental simulation results showed that, compared with several existing studies, our approach can always obtain the highest reward for service vehicles and lowest latency for user vehicles.

Advanced Driver Assistance System Based on IoT V2V and V2I for Vision Enabled Lane Changing with Futuristic Drivability

In conventional modern vehicles, the Internet of Things-based automotive embedded systems are used to collect various data from real-time sensors and store it in the cloud platform to perform visualization and analytics. The proposed work is to implement computer vision-aided vehicle intercommunication V2V (vehicle-to-vehicle) implemented using the Internet of Things for an autonomous vehicle. Computer vision-based driver assistance supports the vehicle to perform efficiently in critical transitions such as lane change or collision avoidance during the autonomous driving mode. In addition to this, the main work emphasizes observing multiple parameters of the In-Vehicle system such as speed, distance covered, idle time, and fuel economy by the electronic control unit are evaluated in this process. Electronic control unit through brake control module, powertrain control module, transmission control module, suspension control module, and battery management system helps to predict the nature of drive-in different terrains and also can suggest effective custom driving modes for advanced driver assistance systems. These features are implemented with the help of the vehicle-to-infrastructure protocol, which collects data through gateway nodes that can be visualized in the IoT data frame. The proposed work involves the process of analyzing and visualizing the driver-influencing factors of a modern vehicle that is in connection with the IoT cloud platform. The custom drive mode suggestion and improvisation had been completed with help of computational analytics that leads to the deployment of an over-the-air update to the vehicle embedded system upgradation for betterment in drivability. These operations are progressed through a cloud server which is the prime factor proposed in this work.

Comparing the sensitivity of the box task combined with the detection response task to the lane change test

The Box Task combined with a Detection Response Task (BT + DRT) is a relatively new and easy-to-use method to assess in-vehicle system demand, consisting of a visual-manual task (BT) and a cognitive task (tactile DRT). Currently, little is known regarding the sensitivity of the BT + DRT for different types and difficulty levels of secondary tasks. Therefore, the present study evaluated the BT + DRT’s sensitivity compared to the Lane Change Test (LCT), which is an ISO-standardized test method. Fifty-two participants engaged in a visual-manual (Surrogate Reference Task) and a cognitive secondary task (counting task) across two levels of difficulty while performing a PC version of the BT + DRT and LCT. The BT parameters, especially the standard deviation of box position and size, were sensitive to visual-manual secondary task demand. Moreover, the results showed that the DRT is a sensitive method to assess cognitive demand, supporting previous findings. In contrast, the mean deviation of the adaptive reference lane used as the standard LCT parameter was not as effective in discriminating between visual-manual and cognitive tasks. Hence, to distinguish between different types and levels of secondary task demand, the BT + DRT is more accurate than the standard LCT version. Future studies should investigate how an increased BT difficulty can further improve the sensitivity to visual-manual and cognitive secondary task demand.

Expected Effect of Adopting the eCall System on Reducing Traffic-Related Deaths and Social Costs

The purpose of this study is to develop an eCall system that can accurately assess the scene of a traffic accident and to analyze the expected effect of the system through integrated demonstrations. The eCall system was designed to collect information requested by emergency response agencies, such as location history immediately before and after an accident, the severity of injuries, and in-vehicle image information. In this study, an eCall in-vehicle system (IVS) was installed in service vehicles, including taxis, fire trucks, and research team members’ cars. Moreover, an eCall center was set up to relay information to emergency response agencies. The results of integrated demonstration tests indicated that the communication success rate between the eCall IVS and the call center was 99.25%, and the mean location information error was 1.2 m. In addition, this study analyzed the amount of time required for an accident victim to arrive at a hospital when the eCall system was implemented, based on 10 years of traffic accident data, and found that he or she would get there approximately 3 min earlier.

Modeling an MMF Connector via Point Radiance Matrix and its Application in Automotive SI-MMF Connection

Recently, multimode fibers and their application, such as in-vehicle systems where many in-line connectors are inserted in short-range transmissions, have garnered considerable research interest. As such applications require extremely high reliability, simulation technology that considers multimode fiber harnessing under unfavorable conditions, such as various fiber lengths and connector misalignments, has been required. Many calculation methods for modal bandwidth and connector loss have been proposed to date; however, the 3D modal power coupling equation at the connection point has not been derived due to the calculation difficulty by linking the near- and far-field patterns. Herein, we propose a ‘point radiance matrix model’ to compute the complex mode-power conversion at the step-index (SI) fiber connectors. Accordingly, we derived new equations for calculating the power distribution at the receiving fiber according to the fiber connector misalignment parameters, which can be used for link level evaluation from light source to detector. Moreover, the rules governing the power transferred from the emitting fiber to the receiving fiber were investigated under generic alignment conditions, including random combinations of longitudinal, lateral, and tilt misalignments. Applications of the proposed matrix method could contribute toward the computation of mode-power conversion at graded-index (GI) fiber connectors in case the point radiance on the emitting fiber end face is found.

Driving Stress Estimation in Physiological Signals Based on Hierarchical Clustering and Multi-View Intact Space Learning

Detecting driver’s statuses is favorable for reducing the incidence of traffic accidents and ensuring driving security. This paper aims to develop an efficient system for driving stress detection under real driving circumstances. Multiple physiological signals, i.e., electrocardiogram (ECG), galvanic skin response (GSR), and respiration (RESP) were collected and multi-modal features were extracted from time, spectral, and wavelet domains. The proposed approaches are motivated by three points: 1) Obvious individual difference affects the transferability of trained models to a new drive. Then, through dissimilarity evaluation and hierarchical clustering, we searched for subgroups of drives that presented relatively consistent feature distributions. Performing cross-drive modeling within each subgroup enables us to identify driver statuses more precisely with less computation cost; 2) fusing the high-dimensional physiological data from multiple views is beneficial to achieve a reliable assessment but brings new challenges for existing techniques. We adopted Multi-view Intact Space Learning (MISL) to integrate rich information from multiple perspectives by constructing a latent intact representation of the data; 3) most of the existing systems are offline. The current study made both offline and online analysis to validate the effectiveness of this research. Experimental results reveal that the proposed approaches can achieve competitive performance to state-of-the-art methods and can be developed into intelligent in-vehicle systems to detect driver’s unfavorable statuses, better adjust their negative affection, and avoid traffic accidents.

Driving With Agents: Investigating the Influences of Anthropomorphism Level and Physicality of Agents on Drivers' Perceived Control, Trust, and Driving Performance.

As technological development is driven by artificial intelligence, many automotive manufacturers have integrated intelligent agents into in-vehicle information systems (IVIS) to create more meaningful interactions. One of the most important decisions in developing agents is how to embody them, because the different ways of embodying agents will significantly affect user perception and performance. This study addressed the issue by investigating the influences of agent embodiments on users in driving contexts. Through a factorial experiment (N = 116), the effects of anthropomorphism level (low vs. high) and physicality (virtual vs. physical presence) on users' trust, perceived control, and driving performance were examined. Results revealed an interaction effect between anthropomorphism level and physicality on both users' perceived control and cognitive trust. Specifically, when encountering high-level anthropomorphized agents, consumers reported lower ratings of trust toward the physically present agent than toward the virtually present one, and this interaction effect was mediated by perceived control. Although no main effects of anthropomorphism level or physicality were found, additional analyses showed that anthropomorphism level significantly improved users' cognitive trust for those unfamiliar with IVIS. No significant differences were found in terms of driving performances. These results indicate the influences of in-vehicle agents' embodiments on drivers' experience.

Engineering Countermeasures for Left Turns at Signalized Intersections: A Review

Left turn crashes can impact the safety of the drivers due to the speed and angle at which they occur. Left turns are specifically reported to affect older drivers more than the other types of crashes. This paper provides a review of the existing engineering countermeasures that have been evaluated to improve driver safety at left turns. Twentyeight studies on left turn signal displays (protected left turns, flashing yellow arrow, and digital countdown timers), intersection geometry (offset left turn lanes, diverging diamond interchange, roundabouts, exit lanes for left turn, left turn bay extension, and contraflow left turn lanes), and driver warning systems (infrastructure warning systems, and in-vehicle warning systems) are reviewed. Eighteen studies were evaluated in the field, nine in laboratory environments, and one online. All countermeasures demonstrated varying levels of effectiveness. We found protected left turns, roundabouts, and warning systems to be the most effective engineering countermeasures. Advantages and disadvantages of each countermeasure and research shortcomings of the evaluation studies are discussed. Review findings may help practitioners and researchers guide more effective countermeasures for left turns for older drivers.

A Multi-Stage, Multi-Feature Machine Learning Approach to Detect Driver Sleepiness in Naturalistic Road Driving Conditions

Driver fatigue is a contributing factor in about 20% of all fatal road crashes worldwide. Countermeasures are urgently needed and one of the most promising and currently available approaches for that are in-vehicle systems for driver fatigue detection. The main objective of this paper is to present a video-based driver sleepiness detection system set up as a two-stage model with (1) a generic deep feature extraction module combined with (2) a personalised sleepiness detection module. The approach was designed and evaluated using data from 13 drivers, collected during naturalistic driving conditions on a motorway in Sweden. Each driver performed one 90-minute driving session during daytime (low sleepiness condition) and one session during night-time (high sleepiness condition). The sleepiness detection model outputs a continuous output representing the Karolinska Sleepiness Scale (KSS) scale from 1–9 or a binary decision as alert (defined as KSS 1–6) or sleepy (defined as KSS 7–9). Continuous output modelling resulted in a mean absolute error (MAE) of 0.54 KSS units. Binary classification of alert or sleepy showed an accuracy of 92% (sensitivity = 91.7%, specificity = 92.3%, F1 score = 90.4%). Without personalisation, the corresponding accuracy was 72%, while a standard fatigue detection PERCLOS-based baseline method reached an accuracy of 68% on the same dataset. The developed real-time sleepiness detection model can be used in the management of sleepiness/fatigue by detecting precursors of severe fatigue, and ultimately reduce sleepiness-related road crashes by alerting drivers before high levels of fatigue are reached.