Browse Topic: Active safety systems

Items (2,179)
With the continuous advancement of vehicle safety technologies, the Automatic Emergency Braking (AEBS) system has become a key component in preventing traffic accidents. However, existing research and testing primarily focus on conventional road conditions, while the system’s performance under extreme environments such as snow and ice remains insufficiently investigated. This paper investigates the challenges encountered by AEBS when addressing Vulnerable Road Users (VRU) in low-friction snowy conditions, analyzing the degradation of sensor performance, uncertainties in decision-making, and actuator limitations arising from reduced friction and low-temperature effects. Furthermore, current regulatory frameworks lack comprehensive test scenarios for icy environments, with existing standards providing only limited assessments under static or simple motion conditions. To address these gaps, this study proposes six representative test scenarios defined by varying snowfall intensity levels, target object types, and vehicle motion states. The evaluation criteria focus on collision avoidance and vehicle stability after braking. The findings contribute to a more systematic and scientifically validated testing methodology for AEBS in winter environments, offering critical insights for future regulatory development and technological advancements in vehicle safety.
Shen, Liangyi, Qi, Yunpeng, Wang, Zhiyu, Xi, Ruixuan, Gao, Zong, Kong, Lingming
Regarding the issue in the 2024 version of C-NCAP’s automatic emergency braking (AEB) test where the scene score rates are highly concentrated at the maximum score, and traditional continuous regression modeling is difficult to effectively depict system risks, this paper transforms the AEB performance assessment problem into a rare failure risk prediction problem, and constructs an AEB failure risk assessment method based on LightGBM and SHAP. Based on 464 test samples, an explanatory variable system is established around operating condition parameters, vehicle configuration parameters, and process temporal sequence parameters, and samples with scene score rates lower than 1 are defined as failure samples. Firstly, the original data undergoes outlier removal, missing value handling, categorical variable encoding, and continuous variable standardization; secondly, a robust LightGBM classification model is constructed to predict AEB deduction/ failure risk; finally, the SHAP method is used to conduct interpretability analysis of the model to identify the key factors affecting AEB system performance. The results show that the established robust model achieves good recognition effects on the test set, with an AUC of 0.871, PR-AUC of 0.527, and F1 of 0.667. The feature importance and SHAP analysis results show that the number of sensors, relative speed, alarm-time TTC, braking-time TTC, overlap rate, and maximum braking deceleration are the key variables affecting AEB failure risk. Among them, higher sensor redundancy and earlier alarm and braking trigger times can significantly reduce system failure risk, while high relative speed, complex collision geometric relationship, and braking intervention lag will significantly increase the failure probability. The proposed method can effectively identify AEB system performance risks under the condition of highly concentrated score rates and rare failure samples, providing a data-driven analysis tool for AEB calibration optimization, sensor configuration design, and active safety performance improvement.
Yuan, Yue, Li, Yanchen, Zhang, Changlu, Li, Zhengwei
In this paper, a safety-space-based hierarchical speed control strategy for high-speed vehicle-pedestrian conflict scenario is proposed. In this study, we call high-speed conflict scenario a transient interaction event in which the vehicle that is in the operation of relatively high speed and the crossing pedestrian may be in the same conflict point in the limited time-space margin. For such risk characterization, the spatiotemporal conflict model is established and the safety-space indicator with the incorporation of the braking demand and the time-to-arrival difference is defined. Then, based on the remaining safety space and the vehicle-pedestrian arrival-time gap at the conflict point, the conflict risk is classified into four levels, and the corresponding hierarchical speed control rules are designed. A VISSIM-based simulation scenario is constructed and the proposed method is compared with No Control and Fixed Deceleration benchmark strategy in terms of the safety, efficiency and smoothness. The results show that, in the simulated scenarios, the proposed strategy can improve the safety-efficiency balance by increasing minimum safety distance, suppressing high risk states, and reducing the unnecessary severe braking compared with the benchmark strategy. The proposed framework provides a preliminary methodological reference for active speed control for high speed vehicle-pedestrian interaction.
Guo, Jing, Wang, Jiaming, Qin, Wei, Shen, Mengyi
Based on standardized C-NCAP active emergency braking (AEB) test data, this paper proposes an integrated performance evaluation method, termed the AEB Performance Score (APS), for time-series test data, aimed at quantifying the active safety performance of AEB systems under different target object types and initial speed conditions. The proposed method integrates key indicators, including avoidance outcome, TTC at AEB activation, maximum deceleration, speed reduction ratio, and stopping margin, and achieves unified evaluation of successful avoidance and collision-failure samples through normalization and hierarchical weighted mapping. Analysis based on real vehicle test data demonstrates that APS can effectively characterize AEB performance boundaries and differences across test scenarios. Correlation analysis further shows that APS is more strongly driven by the speed reduction ratio and stopping margin, indicating that early triggering alone is insufficient to ensure effective collision avoidance. The key to improving AEB performance lies in more accurate deceleration planning and execution consistency. This study provides a new analytical perspective for fine-grained performance evaluation and failure mode identification of AEB systems under standardized test conditions.
Li, Yupeng, Li, Kai, Zhang, Zhiqiang, Geng, Dongliang, Zhang, Shuai, Ma, Wenbo
Adverse rainy and foggy weather is a serious threat to road traffic safety. The scattering and absorption of light by such weather not only limits the field of vision of the driver but also deteriorates critical signals such as brake lights and hazard warning lights into fuzzy and diffused light spots in images. This results in a significant decline in the perception performance of traditional advanced driver assistance system, which increases the risk of rear end collision and other accidents. Under such circumstances, accurately understanding the light signals of preceding vehicles such as braking, turning and hazard warning lights is the most important way to understand the driving intentions of preceding vehicles and carry out active safety warning. To resolve this problem, this paper proposes a light signal recognition method for rainy and foggy scenes based on semantic segmentation and builds a safety-assisted driving model accordingly. A pixel-level annotated dataset including different rainy and foggy scenarios is constructed, and then a semantic segmentation network is trained to accurately detect brake lights and hazard warning lights directly from damaged images. On this basis, the light signal recognition algorithm is deployed on the vehicle end. By setting various conditions on the CarSim simulation platform, the positive significance of the safety-assisted driving model based on light signal recognition proposed in this paper in promoting driving safety in rainy and foggy scenes is proven.
Zeng, Zhixuan, Zhu, Wenfeng, Xue, Ziqi, Kong, Wenkai
To thoroughly mine test data, break down information barriers among automakers, enhance domestic vehicle safety technologies, and reduce road traffic accidents, this paper conducts a three-dimensional analysis of massive test datasets. The study focuses on the 2021 and 2024 versions of the C-NCAP protocols, covering full-overlap, partial-overlap, and newly added scenarios, thereby constructing a multi-level, multi-dimensional data analysis framework. Different versions of test protocols inevitably exhibit both continuity and updates over time. By analyzing the test results, significant performance differences among various vehicle models were identified in scenarios involving high speeds, turning maneuvers, and intersections. This work provides a comprehensive demonstration of diverse vehicle operating conditions and holds substantial research value. This paper fully demonstrates the active safety performance of different vehicle models and offers new research insights for the updating and iteration of the C-NCAP protocols.
Duan, Benyi, Fu, Wang, Song, Jia, Zhang, Cheng, Ma, Wenbo, Wang, Botong
As global vehicle numbers rise, traffic accidents pose significant threats to life and property, with approximately 1.3 million casualties annually. To mitigate accidents, automotive Original Equipment Manufacturers and suppliers are prioritizing active safety technologies like the automatic emergency steering (AES) system, which proactively intervenes in vehicle dynamics to avoid collisions. The AES system, as an important component of active safety, lacks systematic testing and evaluation methods. We aim to conduct a more comprehensive and scientific pre-evaluation of the AES system, which realizes its functions through onboard sensors, decision control systems, and actuators. The environmental awareness system uses sensors such as a camera, millimeter wave radar, and laser radar to collect surrounding information. Based on the perceived data, the decision control system generates the optimal steering strategy through hazard identification and risk assessment. This study addresses these gaps by analyzing AES performance indices and existing regulations (e.g., UN R79, ISO 23375, IIHS, E-NCAP) to design a comprehensive test framework. Four high-coverage scenarios were developed: longitudinal scooter conflicts, vehicle cut-ins, pedestrian crossings, static cone bucket, and misoperation validation, ensuring evaluation under diverse conditions. It has important practical significance for improving China’s automotive safety technology level and reducing road traffic accident casualties. With the rapid development of autonomous driving technology, this research achievement will also provide technical support for the development of higher-level intelligent driving systems.
Shen, Liangyi, Gao, Zong, Kong, Lingming, Wang, Zhiyu, Zhang, Xiao
Aiming at the problems that existing energy-saving technologies for new energy vehicles highly rely on high-cost, high-precision sensors such as radars and cameras, and suffer from high implementation costs, this paper studies vehicle operation scene recognition methods for energy-saving driving of new energy vehicles, aiming to develop a low-cost scene recognition model and provide corresponding technical support for energy-saving driving of such vehicles. Firstly, actual road driving data of new energy vehicles with comprehensive scene coverage are collected via real-vehicle tests, with the total test mileage exceeding 126,000 km. Secondly, a data-driven recognition method for typical vehicle operation scenes combining Principal Component Analysis (PCA) and the K-means clustering algorithm is proposed, which classifies three typical operation scenes: congested, urban free-flow, and suburban scenes. Finally, a set of scene-specific driving assistance control strategies is designed, and the potential energy-saving effects of the strategies based on vehicle operation scene recognition are investigated based on a mature powertrain energy consumption simulation model for new energy vehicles. The research shows that under the China Light-Duty Vehicle Test Cycle (CLTC), the energy consumption of new energy vehicles can be reduced by approximately 4.5% through anticipatory perception of vehicle operation scenes and adaptive switching of scene-specific energy-saving control strategies. Moreover, the proposed technology can realize accurate scene recognition without relying on high-cost, high-precision sensors such as radars and cameras, and it can meet the application requirements of low-cost new energy vehicles, including logistics vehicles.
Liu, Yu, Guo, Yijing, Zhang, Hao, Yu, Hanzhengnan, Yuan, Xiaohui
As auto-driving technologies advance, the driver monitoring system (DMS) has shifted from a premium option to a standard feature of passenger vehicles. However, the compatibility and coordination between the DMS and the main active safety functions (ASF) of vehicles, such as automatic emergency braking (AEB), forward collision warning (FCW), lane departure warning (LDW) and lane keeping assistance (LKA), have been rarely studied. To fill this research gap, this study first categorized unsafe driving risks into three levels, which serve as a quantitative basis for DMS activation; Then, a novel collaboration model between DMS and ASF is constructed, which outputs dynamic ASF adjustment strategies based on the driver’s status. A multi-dimensional quantitative evaluation indicator system is then constructed to assess the effectiveness of the proposed model and the corresponding strategies. Moreover, controlled tests are performed to verify the model’s effectiveness, and validation tests are performed to confirm that it cannot only enable adaptive optimization of the warning timing and intervention intensity of ASFs, but also balance safety redundancy and driving experience. This study realizes coordination between the DMS and vehicles’ ASFs and provides a reference for optimized design of safety functions of intelligent vehicle safety systems.
Liu, Jiajie, Yang, Guhan, Zhang, Cheng, Ma, Wenbo, Wang, Botong
In response to the lack of test scenarios and evaluation methods for multi-object pedestrian crossing AEB(Advanced/Automatic Emergency Braking) systems, this paper proposes a multi-object pedestrian crossing AEB test scenario and an associated evaluation method. By analysing the road safety issues arising from pedestrians crossing the road and collision-avoidance technologies designed for vulnerable road users such as pedestrians, and by integrating existing regulations and testing and evaluation frameworks for pedestrian crossing AEB systems, a comprehensive testing and evaluation framework for multi-object pedestrian crossing AEB systems has been established.
Li, Yanchen, Wang, Zhiyu, Ji, Zhao
In recent years, Reinforcement Learning (RL) has been increasingly applied to the collision avoidance systems of Unmanned Aerial Vehicles (UAVs) due to its high performance in complex tasks and relatively low online computational cost. Despite these advantages, the lack of interpretability in RL and decision-making processes poses significant challenges to safety verification, which severely limits the practical deployment of RL-based autonomous flight systems in safety-critical scenarios. Ensuring safety during both training and testing remains the key for real-world applications. To address this issue, a Run Time Assurance (RTA) filter framework for a UAV collision avoidance system based on RL is proposed. The framework continuously monitors the avoidance actions generated by RL and ensures that unsafe actions can be identified and corrected in real time. The framework provides security while retaining the performance advantages of RL. First, a discrete kinematics model of the UAV and a collision avoidance environment for a three-dimensional multi-intruder UAV are established. The Proximal Policy Optimization (PPO) algorithm is used to train the UAV avoidance policy. Secondly, the RTA filter with safe constraints is designed based on this model. To reduce the position error between the discrete kinematics model and the continuous dynamics model of the UAV, optimization is performed using the Sequential Least Squares Programming Method (SLSQP). The RTA is applied to the training and testing phases of UAV collision avoidance systems to monitor and correct unsafe actions online. Finally, three training and testing methods are set: no RTA, RTA without punishment, and RTA with punishment. Comparing different relaxation coefficients allows for analyzing their impact on filter performance. Simulation results demonstrate that the RTA effectively enhances safety throughout training and testing. Compared with the no RTA method, the collision rate is reduced by 7.04%, and the success rate is increased by 2.68%. In extreme collision avoidance scenarios, the average collision rate obtained by different reinforcement learning algorithms is reduced by 21.46%, and the average success rate is increased by 15.54%.
Ma, Zan, Yan, Liqin, Bai, Jie, Liu, Yubin, Sun, Shuguang
Automotive radar systems do not authenticate the signals they receive, allowing an external transmitter to create phantom objects that can trigger automatic emergency braking (AEB). This paper quantifies the threat using real radar measurements from the public RadarScenes dataset and evaluates a software countermeasure that requires no hardware modification. Computer-generated phantom radar returns were added to 200 motorway sequences and adjusted so the OS-CFAR detector accepted them as real objects. In the undefended system, radar detections were tracked as obstacles and braking was initiated when an object appeared within 30 m, resulting in unnecessary braking in 87.0 percent of attacked sequences. The proposed two-layer filter combines kinematic plausibility gating with camera-radar cross-validation. Layer 1 checks track history and physical movement over three consecutive frames, while Layer 2 reduces confidence when a clear-weather radar detection lacks camera confirmation. With both layers active, false braking fell to 3.5 percent, with no genuine detections missed across 200 genuine sequences. SHAP analysis of a Random Forest classifier identified kinematic implausibility as responsible for 62.3 percent of correct suppressions and produced a machine-readable evidence record to support cybersecurity documentation under ISO/SAE 21434 and safety documentation under ISO 21448. These results apply to an offline replay experiment using synthetic injections. Validation with real attack hardware, synchronized camera data, and vehicle testing remains future work.
Sehgal, Vishal, Sekaran, Nikhil
As automotive architectures evolve toward Software Defined Vehicles (SDVs) and high levels of driving automation, traditional safety analysis techniques rooted in hardware reliability require significant adaptation. The Classical Event Tree Analysis (ETA), a cornerstone of ISO 26262 functional safety compliance, relies on static, binary branching logic where each function either fails or succeeds. While highly interpretable and effective for modelling random hardware failures, classical ETA struggles to capture the dynamic, non-linear interactions, temporal degradation, and complex human-machine interface (HMI) dependencies that are characteristic of modern Advanced Driver Assistance Systems (ADAS) and autonomous vehicles. This paper proposes State Dependent Event Tree Analysis (SD-ETA), a methodological extension designed to integrate discrete state vectors and temporal dependencies into the classical ETA framework. SD-ETA assigns probabilities to the branches for the dynamic nonlinear interactions, temporal degradation, and complex human-machine interface (HMI) dependencies instead of binary values, thereby allowing the system state (comprising vehicle physics, environmental conditions, and driver attention metrics) to dynamically update based on the outcomes of the preceding nodes. This approach retains the visual hierarchy and regulatory familiarity of standard event trees while significantly enhancing model fidelity for Safety of the Intended Functionality (SOTIF) scenarios governed by ISO 21448. To demonstrate the necessity and utility of this methodological evolution, this paper presents a comparative case study analysing a system operating under degraded environmental conditions. The quantitative application reveals that while classical ETA predicted a relatively low collision risk by treating sensor performance as a static binary state, the dynamic SD-ETA simulation exposed a critical hidden safety gap, calculating a higher collision probability due to compounded latency and degraded physical operating parameters. Ultimately, SD-ETA equips safety practitioners and automotive regulators with a modernized, time-aware framework to accurately evaluate the complex, interacting risks of autonomous systems without sacrificing the transparency required for rigorous safety certification.
Sharma, Dhiraj Mohan, Mutagi, Megha, Reddy, Kakarla Subba
Children dying in hot cars is not a new problem. Since 1998, more than 1,000 children have died this way in the United States alone — roughly 37 every year. That number has barely been reduced, which tells something alarming: the solutions built so far aren't solving it completely. Door-based reminders are a good example of why. They trigger when a rear door opens, but opening a door doesn't confirm a child was left behind. Single-sensor setups have their own gaps too; no single sensor handles every real-world scenario reliably on its own. Multi-sensor systems get closer, but most use rigid threshold-based rules that fall apart the moment conditions deviate — a sleeping infant barely moves, a child under a blanket reads completely differently on a thermal sensor. This paper proposes Smart-ROPS (Smart Rear Occupant Protection System), an embedded ML architecture that targets these specific failure modes without needing expensive hardware or complex software. Four sensor types feed the system: millimeter-wave radar, CO₂, thermal infrared, and ultrasonic. Data from these passes through three layers — the Perception Layer handles initial sensor fusion, the Processing Layer runs a decision tree as the primary classifier, and when confidence is low, a shallow neural network steps in for ambiguous cases. If a risk is confirmed, the Intervention Layer manages the response: starting with a smartphone alert and cabin ventilation, escalating to emergency services if needed. This is presented as an architectural design contribution — hardware prototyping and real-world testing are left for future work.
Shrimal, Harsh
Autonomous Emergency Braking (AEB) is a safety-critical Advanced Driver Assistance System now mandated for commercial vehicles in India.to reduce crash frequency and severity. This paper presents a structured, simulation-based V&V approach for camera-based AEB that integrates TTC braking evaluation, Forward Collision Warning assessment, and real-time object detection validation. High-fidelity tools Hexagon VTD & ROD, and MATLAB Roadrunner are employed to generate photorealistic, India-specific scenarios capable of modelling detailed road geometry, marked and unmarked lanes, mixed traffic behavior, and diverse environmental and weather conditions. Scenario-driven testing enables frame-accurate, end-to-end assessment of detection, decision-making, and actuation without requiring access to internal algorithms (black-box validation). Across representative scenarios, the test campaign achieved up to 90% accuracy in detecting relevant targets. This validation aligns with AIS & MoRTH guidance, and adheres to UN Regulation (UN R131), while incorporating applicable ISO 26262 functional safety principles. Requirement traceability matrices are used to evaluate system behavior under sensor degradation, misclassification, and other critical edge-case conditions prevalent in Indian driving environments. The proposed methodology enables consistent, repeatable, and regulation aligned validation of camera based AEB systems, supporting safer and more reliable commercial vehicle deployment in India.
S., Nagabhushana N., P., Gururaaj, M. L., Sankar
In autonomous driving and advanced driver-assistance systems (ADAS), radar sensors play a crucial role in detecting and tracking objects under diverse environmental conditions [1, 2, 3]. One persistent challenge is the accurate representation of long vehicles, such as trucks and buses, whose extended radar signatures often result in a wide spatial spread of detections with similar radial velocities, making association difficult due to gating limitations. The GRID component processes raw radar measurements to generate occupancy grids and extract candidate targets, including stationary environmental structures [4, 5]. In the case of stationary long vehicles, multiple stationary objects may overlap with grid-based environment targets, allowing them to be consolidated into a single tracked object. This paper presents a novel approach for radar object merging and retention decision-making using fuzzy logic to distinguish between true moving and stationary objects in the presence of GRID targets and object association mechanisms. The association modules link radar measurements or detections, when available, with existing tracks to maintain temporal consistency, enabling the formation of radar objects represented by track points. The proposed method fuses radar objects based on their radial velocity and spatial overlap, while a fuzzy logic–based decision engine evaluates object persistence. The approach improves object continuity and reduces false positives in driving functions such as Blind Spot Detection (BSD), Emergency Brake Assist (EBA), and Adaptive Cruise Control (ACC). Simulation and real-world validation results demonstrate significant improvements in object tracking accuracy, particularly for long vehicles in urban environments.
Mohanty, Sanket, Sarkar, Sanglap
The application of Advanced Driver Assistance Systems (ADAS) in commercial vehicles marks significant challenges in mechanical design and digital validation of externally mounted sensor systems. Blind Spot Assist (BSA), a radar-based ADAS feature, depends on accurate sensor alignment and stable bracket mounting to ensure reliable detection of objects within the vehicle’s blind spots. the exposure to dynamic loading environments, BSA radar brackets must be designed to withstand operational vibrations and also should comply with international automotive standards, including ISO 16750-3 [1, 7], which outlines test procedures of mechanical vibration and shock. This paper describes a digital approach for the development of BSA radar brackets, incorporating system-level design thinking, flexible-body simulation, and vibration analysis. The method is combination of physical behavior with simulation techniques to study bracket behavior for dynamic loads and identifying probable sources of structure resonance and the interaction of the system with neighboring systems or components. A multi-dimensional design process is highlighted, where multiple iterative modeling, cross-functional teams brainstorming, and virtual validation enables the efficient exploring of alternate designs. The framework supports countermeasure studies such as geometry optimization, boundary condition refining, and system decoupling, with digital development. This approach is applicable for a wide range of ADAS sensor mounts challenges faced in heavy-duty vehicles, where the packaging constraints, harsh operating conditions, and NVH (Noise, Vibration, Harshness) requirements must be considered. This paper shows a replicable model for engineering teams to improve sensor stability, reduce physical testing and simulation cycles, and ensure compliance with vibration standards from the design evolved from simulations.
Kenchaplara, Vinay Kumar, Hubballimath, Nandesh, Sadekar, Rahul
Overtaking another vehicle on a two-lane road can turn dangerous within a few seconds, especially when the gap to an oncoming vehicle is short or that vehicle is moving fast. Many driver-assistance systems rely on a simple, fast rule called the Overtake Safety Margin formula to decide whether an overtaking maneuver is safe, without running a full traffic simulation. This study checks the formula's reliability using a physics-based simulator that models three vehicles: the car overtaking, the car being overtaken, and the oncoming car. One million generated overtaking scenarios were evaluated using both the OSM formula and the kinematic simulator to identify cases where the formula classified a maneuver as SAFE while the simulator predicted an UNSAFE outcome. These mismatches, called blind spots, represented 3.98% of the evaluated scenarios and were concentrated in high oncoming-speed, short-gap situations, pointing to a specific weakness rather than random error. To catch these blind spots automatically, a small and efficient machine-learning model, a micro-transformer, was trained on the simulator's results. It reached 96.3% overall accuracy and correctly flagged 86.9% of the formula's blind spots, outperforming LightGBM, which detected 73.2%. The model's confidence scores were well calibrated, so they can be trusted to trigger an extra safety check when needed. An uncertainty-based search then used these confidence scores to find risky scenarios efficiently: 187 of the 500 most uncertain cases turned out to be unsafe, about 9.4 times more effective than random selection. Finally, the 30 highest-priority cases were replayed in the CARLA driving simulator, and 27 were confirmed to reproduce unsafe outcomes. Together, these results show that pairing a fast rule-based formula with a calibrated learning-based monitor can catch dangerous situations the formula alone would miss, supporting the safety documentation required for automotive safety-of-the-intended-functionality standards.
Sekaran, Nikhil, Sehgal, Vishal
Bicycle computers and apps record, at minimum, positional data over time, and these data are commonly used in accident reconstruction to understand the behavior of the bicycle and rider prior to an incident in question. These positional data are obtained using the Global Positioning System (GPS), and while their absolute positional accuracy has been the subject of prior research, their accuracy at detecting and reporting particular movements is less studied. To improve the accident reconstruction industry’s understanding of these devices’ performance, this research aims to statistically quantify the temporal and positional accuracy of these devices reporting the onset of a lateral deviation or lane change. Controlled testing was performed and recorded with several commercially available bicycle GPS computers and apps, which were compared to a RaceLogic VBox 3i ADAS with Real-Time Kinematics (RTK) corrections from a RaceLogic Base Station. Three separate test bouts were performed, with each test bout consisting of 30 or 32 repeats of three different lateral deviation maneuvers. The bicycle GPS computers were individually synchronized to the RaceLogic data by offsetting their time to minimize the mean-square positional error across the entire test bout, which enabled calculation of the 50th percentile and 95th percentile absolute positional errors for each device. A custom script was then used to programmatically detect the start of each lateral deviation, and then, confidence intervals were calculated to estimate the probability of each GPS device reporting the start of the lateral deviation with zero lead or lag, with 1 s of lag, or with 0 or 1 s of lag based on the relative timestamps and positional data. All three of the tested bicycle GPS computers had a probability of at least 0.5 of reporting the onset of sharp lateral movements with zero lead or lag based on the time data, while only two of the devices maintained a similarly high probability for the position-based data. The iPhone 17 Pro had a probability greater than 0.6 of detecting the onset of both gradual and sharp lateral movements with 1 s of lag for both the time-based and position-based data. And across all lateral movement types, all devices had a probability of at least 0.6 of reporting the onset of lateral movement with either 0 or 1 s of lag.
Sweet, David Michael, Bretting, Gerald, Wilhelm, Chris, O’Brien, Nathan
Modern automotive platforms must serve multiple market segments amid rapid technological change and stringent regulation, making early concept selection a multi-criteria, product-line problem. This paper presents a repeatable workflow that integrates Model-Based Product Line Engineering (MBPLE), Multi-Criteria Decision Analysis (MCDA), and enterprise visualization. A Systems Modeling Language (SysML) 150% vehicle architecture in Cameo captures powertrain, chassis, and Advanced Driver Assistance Systems (ADAS) variability plus baseline and segment-specific requirements. Parametric models compute vehicle-level attributes (e.g., cost, mass, braking capacity, detection performance, Technology Readiness Level (TRL)) and enforce segment limits. A multi-attribute value theory (MAVT) framework model is implemented as constraint blocks to normalize attributes and aggregate stakeholder-elicited weights into an overall score per configuration. Cameo Trade Study sweeps the design space, exports a configuration–criteria dataset, and Power Business Intelligence (BI) dashboards enable interactive cost–value views, requirement compliance, and shortlist comparisons. Selected concepts are fed back into Cameo as feature configurations, improving traceability, scalability across product lines, and alignment between engineering models and management decisions.
Kliczinski, Gary, Pykor, Ryan
Advanced Driver Assistance Systems (ADAS) are evolving beyond onboard perception. The ability to dynamically map and share temporary road hazards is important for connected and autonomous driving, but authenticating the data is critical. An in-vehicle hazard recognition layer performs real-time video analysis and geotagging on embedded platforms. Real-time video is analyzed for road hazards—such as potholes, construction zones, waterlogging, fallen trees, roadside accidents, and debris—using onboard cameras and lightweight computer vision models. This metadata is sent to a cloud-based aggregation layer to validate hazard reports. It is of paramount importance to validate hazard reports originating from diverse sources, regardless of their accuracy. This paper presents a mathematical approach to determine an overall Hazard Confidence Score (HCS) based on data received from diverse sources. A unique hazard authentication model is introduced to quantify the credibility of each hazard report using six validation metrics.
Bose, Souvik
This document describes [motor] vehicle driving automation systems that perform part or all of the dynamic driving task (DDT) on a sustained basis. It provides a taxonomy with detailed definitions for levels of driving automation, ranging from no driving automation (Level 0) to automated driving under all conditions in which humans can drive, with human driving not needed (Level 5), in the context of [motor] vehicles (hereafter also referred to as “vehicle” or “vehicles”) and their operation on roadways: Level 0: No driving automation Level 1: Driver support for steering OR speed, with continual driver supervision necessary and driver intervention when needed Level 2: Driver support for steering AND speed, with continual driver supervision necessary and driver intervention when needed Level 3: Automated driving under defined conditions, with human driving needed following an alert or evident vehicle malfunction Level 4: Automated driving under defined conditions, with human driving not needed to mitigate risk Level 5: Automated driving under all conditions in which humans can drive, with human driving not needed. The simple level descriptors have been changed to improve understanding of the differences among levels, but these are NOT the definitions of the levels of driving automation. See the definitions of each automation level in Sections 4 and 5 for explanation of these changes. These level definitions, along with additional supporting terms and definitions provided herein, can be used to describe the full range of driving automation features equipped on [motor] vehicles in a functionally consistent and coherent manner. “On-road” refers to publicly accessible roadways (including parking areas and private campuses that permit public access) that collectively serve all road users, including cyclists, pedestrians, and users of vehicles with and without driving automation features. The levels apply to the driving automation feature(s) that are engaged in any given instance of on-road operation of an equipped vehicle. As such, although a given vehicle may be equipped with a driving automation system that is capable of delivering multiple driving automation features that perform at different levels, the level of driving automation exhibited in any given instance is determined by the feature(s) that are engaged. This document also refers to three primary actors in driving: the (human) user, the driving automation system, and other vehicle systems and components. These other vehicle systems and components (or the vehicle in general terms) do not include the driving automation system in this model, even though as a practical matter a driving automation system may actually share hardware and software components with other vehicle systems, such as a processing module(s) or operating code. The levels of driving automation are defined by reference to the specific role played by each of the three primary actors in performance of the DDT and/or DDT fallback. “Role” in this context refers to the expected role of a given primary actor, based on the design of the driving automation system in question and not necessarily to the actual performance of a given primary actor. For example, a driver who fails to monitor the roadway during engagement of a Level 1 adaptive cruise control (ACC) system still has the role of driver, even while they are neglecting it. Active safety systems, such as electronic stability control (ESC) and automatic emergency braking (AEB), and certain types of driver assistance systems, such as lane keeping assistance (LKA), are excluded from the scope of this driving automation taxonomy because they do not perform part or all of the DDT on a sustained basis, but rather provide momentary intervention during potentially hazardous situations. Due to the momentary nature of the actions of active safety systems, their intervention does not change or eliminate the role of the driver in performing part or all of the DDT, and thus are not considered to be driving automation, even though they perform automated functions. In addition, systems that inform, alert, or warn the driver about hazards in the driving environment are also outside the scope of this driving automation taxonomy, as they neither automate part or all of the DDT, nor change the driver’s role in performance of the DDT (see 8.13). It should be noted, however, that crash avoidance features, including intervention-type active safety systems, may be included in vehicles equipped with driving automation systems at any level. For automated driving system (ADS) features (i.e., Levels 3 to 5) that perform the entire DDT, crash mitigation and avoidance capability is part of ADS functionality (see also 8.13). Note that this document provides a taxonomy and definitions and is not a safety standard. The document is not intended to provide guidance for safe vehicle operation by the driving automation system.
On-Road Automated Driving (ORAD) Committee
With the continuous development of autonomous driving technology, vehicle collision warning systems are playing an increasingly important role in this field, promoting the progress and improvement of the whole autonomous driving field. But existing methods have low detection rates and unstable multi-target tracking performance. In particular, the estimation of relative motion parameters is still inaccurate due to the loss of direction information when relative speed is described as a scalar quantity. These limits produce easy collisions in the judgment of the car in a dangerous traffic environment. To solve these problems, a vehicle collision warning algorithm based on YOLOv8 and DeepSORT is proposed in this paper. YOLOv8 is introduced to detect the vehicle precisely, and DeepSORT is used to enhance multi-target vehicle tracking. The geometric principles of monocular vision are applied to extract key motion parameters such as distance and direction-signed relative speed. A classification logic is designed to distinguish between positive and negative relative velocities, enabling more accurate judgment of collision risk levels. In order to further enhance the reliability of the system, a four-stage cascaded false-alarm suppression mechanism is proposed. By adding velocity direction validation, distance validity checks, adaptive confidence thresholds, and a temporal consistency verification mechanism, the false alarm rate is reduced, and the proposed approach can realize direction- aware velocity estimation without requiring additional sensors and can be easily integrated into the existing YOLOv8 perception system.
Qin, Xiaoyu, Li, Wei
An adaptive performance-enhanced path planning algorithm is proposed for unmanned surface vehicle (USV) to improve their responsiveness in dynamic maritime environments. The improved ant colony (ACO) algorithm incorporates a pheromone penalty mechanism and path smoothing to enhance search efficiency and path smoothness by removing redundant nodes and reducing excessive turning. Additionally, the dynamic window approach (DWA) is enhanced through three key modifications: optimizing overshoot, enhancing selection efficiency in candidate path, and adaptively adjusting evaluation function weights. These improvements improve the accuracy of planning and avoidance ability. Comparative analysis based on simulation data indicates that the proposed method yields a measurable improvement in path quality—characterized by reduced travel length and enhanced collision avoidance—leading to more robust navigation performance in complex marine transportation scenarios.
Sun, Jiamian, Li, Weifeng
Typical maritime monitoring scenarios are usually constrained by factors such as multi-scale ship density, frequent motion overlap, and limited viewing angle of shore-based cameras. These challenges often lead to trajectory interruptions and identity mismatches in target detection and multi-target tracking tasks. In order to solve these problems, this study proposes a ship occlusion detection and tracking method based on the improved YOLOv8 model and further integrates an automatic identification system (AIS) trajectory reasoning. The method builds a unified perception framework with enhanced detection architecture, multi-source data fusion, and behavioral reasoning capabilities. First, in the target detection module, the improved SEConv structure is introduced into the YOLOv8 trunk network to address challenges caused by small-scale variations and severe occlusion in maritime scenes. The ReLU activation function in SEConv is replaced by the Swish activation function to enhance the nonlinear feature representation. In addition, the optimized SEConv is embedded in the C2f structure, and the convolutional block attention module (CBAM) attention mechanism is introduced to enhance the sensitivity of the model to the occlusion area. Next, for multi-target tracking, ByteTrack is used as the basic tracking framework. AIS trajectory data is introduced as auxiliary input to compensate for trajectory losses caused by occlusion. Finally, experimental results on the SeaShips public dataset and the self-built occlusion reference dataset show that the improved YOLOv8 detector achieves stable mAP gains in mild, moderate, and severe occlusion scenarios. The AIS enhanced tracking system improves the multi-target tracking accuracy (MOTA) and identification F1 score (IDF1) by about 6.3% and 8.1%, respectively, and the average occlusion reconstruction error is controlled within 1.4 seconds. The proposed method effectively enhances the perception ability of ships in complex occlusion environments and verifies the feasibility and superiority of the strategy of combining visual detection with AIS data assistance.
Guan, Keping, Chen, Miao, Zhou, Yue
Research on automatic emergency braking (AEB) control algorithms for heavy vehicles is relatively limited. Compared with passenger cars, heavy vehicle AEB algorithms must accommodate both unloaded and fully loaded conditions, with the latter posing higher demands. This study compares two distinct AEB control strategies: the time-to-collision (TTC) algorithm and the professional driver fitted (PDF) algorithm. Using simulation analyses under three regulatory-recommended scenarios—stationary lead vehicle, slow-moving lead vehicle, and decelerating lead vehicle—the results indicate that the PDF-based control system better adapts to both unloaded and fully loaded conditions. It demonstrates significant improvements in braking performance and robustness compared to the TTC-based system. For an unloaded vehicle equipped with the PDF–AEB control system (5500 kg), the final gap to the lead vehicle is the longest (11.5 m) under the scenario of a stationary lead vehicle with an initial ego vehicle speed of 80 km/h and the shortest (3.1 m) under the scenario of a lead vehicle with an initial speed of 50 km/h braking at 0.4 g. For a fully loaded vehicle (12,500 kg), the corresponding final gaps to the lead vehicle are 11.2 m and 2.5 m, respectively.
Lai, Fei, Huang, Chaoqun
The gathered journalists cover their ears as a loud hydraulic ram begins to load and a warning alarm sounds. Less than a minute later… FOOSH… BAM! A 9.9-pound model of a human head is launched into the front-end structure of a test vehicle, deforming the hood and sending the head bouncing to the ground. The violence of the moment, which is part of pedestrian safety research, is a stark reminder of the seriousness of all the testing that happens at Toyota's North American Research and Development Headquarters in York Township, Michigan. The $187-million (in 2007 dollars) facility includes 180,000 sq.ft. of research and testing space.
Clonts, Chris
Driver’s distraction and fatigue are among the major contributing factors of traffic accidents. This study presents a methodology to identify driver’s distraction using a refined You Only Look Once (YOLO) model, denoted as YOLOv11.To address the inconsistent performance of earlier versions of YOLO, especially with regard to lack of systematic evaluations, this study proposes an improved YOLOv11 model. A mixed local channel attention (MLCA) module is further introduced to enhance small object feature extractions considering the use of Wise-Intersection over Union (IoU) v3 loss function to improve localization accuracy and training stability. Experiments demonstrated that this model outperforms competing models across all metrics, achieving 99.13% mAP at 0.5 and 82.54% mAP at 0.5:0.95, while also achieving minimal bounding box loss. The proposed model demonstrated higher accuracy and robustness, making it suitable for real-world driver monitoring system (DMS) deployments.
Ma, Bao, Taghavifar, Hamid, Fu, Zhijun, Karangwa, Jules, Rakheja, Subhash
As tractor-trailers are essential to global logistics, their roll stability during emergency maneuvers is a critical safety concern. This paper presents a novel delay-compensated active roll control strategy for tractor-trailers using a two-dimensional piston pump electro-hydrostatic actuator (EHA). Unlike existing advanced strategies that assume ideal actuator behavior, this approach specifically targets the inherent response delay in high-tonnage applications. A detailed EHA model, including pump flow characteristics and hydraulic mechanics, was developed and validated through step response experiments. A seven-degree-of-freedom vehicle dynamics model and a model predictive controller were also constructed to compute the required anti-roll moment under emergency driving conditions. In order to address the EHA actuator’s response delay, a delay feedforward controller (DFC) was designed, integrating acceleration feedforward, feedback regulation, and delay disturbance estimation. TruckSim–Simulink co-simulations under double lane-change (DLC) maneuvers at 40 km/h, 60 km/h, and 80 km/h show that DFC improves displacement tracking and reduces peak trailer roll angle by up to 15% compared to a velocity-feedforward proportional-integral-derivative (VFPID) controller. It also enhances control efficiency, as evidenced by lower average motor speeds and pressure response of EHA. The system demonstrates high power-to-weight ratio and efficient tracking capabilities under dynamic conditions. Although active control provides limited benefit at low speeds, the proposed strategy effectively improves roll stability and driving safety under dynamic conditions.
Chen, Lijie, Yin, Yuming, Zeng, Yuhang, Ruan, Jian, Li, Hangqi, Sun, Peng
This specification covers a corrosion- and heat-resistant steel in the form of wire.
AMS F Corrosion and Heat Resistant Alloys Committee
Sonatus and Omdia surveyed over 500 auto executives for their thoughts on SDVs, and the responses revealed that - once again - the future is unwritten. Data monetization as a top priority dropped from 51% to 44% (YoY) as automakers focus more on using data internally, with ADAS and product improvements rising six points to 41%. We spoke with Sonatus CMO John Heinlein about the survey for a recent episode of the SAE Automotive Engineering Podcast. You can listen via the links at the bottom of this Q&A, which is an edited version of our full discussion.
Blanco, Sebastian
Advanced Driver Assistance Systems (ADAS) are increasingly integrated into heavy-duty commercial vehicles to improve road safety, mitigate accident severity, and enhance operational efficiency. In the context of braking systems, however, a significant gap remains between system calibration practices and real-world operating conditions, where most evaluations and validations of ADAS and braking performance are conducted under nominal or standardized load assumptions, which fail to represent the wide variability of payload magnitude and distribution typically observed in trucks, semi- trailers, and buses. Variations in vehicle mass and load distribution directly affect braking efficiency, axle load transfer, center of gravity (CG) position, and stability limits, posing critical challenges to both conventional brake systems and brake-related assistance functions. This paper presents an exploratory qualitative study based on a systematic literature review addressing the influence of variable loading on braking performance and the operation of ADAS in commercial vehicles. Scientific publications, experimental investigations, and technical reports from both academia and industry were thoroughly analyzed, with emphasis on service brake efficiency, load-dependent braking behavior, rollover propensity, and the performance of the ADAS systems. The reviewed studies demonstrate that longitudinal, lateral, and vertical CG displacements significantly modify braking force distribution, actuator effectiveness, and stopping distances, particularly under emergency braking and downhill driving conditions. Improper load distribution was consistently associated with reduced braking margins and increased instability risk. The findings further indicate that integrating load-aware strategies into braking control and ADAS calibration can improve braking consistency, reduce component stress, and enhance overall vehicle safety. As a contribution, this work emphasizes the need to incorporate variable load conditions into braking system evaluation, ADAS development, and certification procedures for heavy commercial vehicles, ensuring robust and reliable performance under real operating conditions.
Rubbo, Bruno Tiago, Dacol, Franco De Bastiani, Do Nascimento, Vagner
The implementation of ADAS in buses represents both a significant opportunity and a complex challenge for the future of urban mobility. While ADAS technologies such as lane departure warning, adaptive cruise control, blind spot detection, and autonomous emergency braking have been widely adopted in passenger cars and trucks, their integration into buses has been slower due to unique operational and safety concerns. This paper provides a broad overview of the advantages and obstacles associated with ADAS deployment in public transport vehicles, with particular emphasis on passenger safety, regulatory frameworks, and operational efficiency. Key barriers include the vulnerability of standing passengers during sudden braking events, the unpredictability of pedestrians and cyclists in dense urban environments, and the economic constraints faced by bus operators. At the same time, regulatory initiatives such as Transport for London’s Bus Safety Standard, the European Union’s General Safety Regulation, and Brazil’s MOVER program are driving the gradual adoption of these systems. The benefits of ADAS in buses extend beyond accident reduction, encompassing improved driver ergonomics, reduced fatigue, lower maintenance costs, and enhanced passenger comfort. Case studies from Europe, Brazil, and Asia highlight both the safety potential and the reluctance of drivers to fully embrace these technologies, often due to knowledge gaps and perceived inconvenience. The analysis underscores that successful implementation requires not only technological adaptation but also comprehensive driver training, infrastructure readiness, and public policy support. Ultimately, ADAS in buses should be understood as a transitional step toward autonomous mobility, offering immediate safety gains while reshaping the paradigm of urban transport.
Marcon, Ederson, Michelon, Gabriel, do Nascimento, Vagner
This paper proposes a nonlinear and robust State-Dependent Riccati Equation (SDRE) combined with H∞ control architecture for brake- by-wire systems, specifically designed to handle severe tire-road friction variations and μ-split scenarios. The primary objective is to maximize deceleration capabilities while rigorously maintaining yaw stability, trajectory tracking, and passenger comfort through jerk limitation. Situated within the domain of active safety, this research addresses robustness against real-world uncertainties by utilizing a high-fidelity 14-degree-of-freedom vehicle model that accounts for longitudinal, lateral, and yaw dynamics, suspension-induced pitch and roll effects, and nonlinear tire behavior with explicit load transfer. To ensure near-optimal slip tracking under variable surface conditions, the system employs online friction estimation via Extended and Unscented Kalman Filters (EKF/UKF) fusing wheel and IMU data to adaptively adjust slip targets. The control strategy is bifurcated: the SDRE component manages dominant nonlinearities through state-dependent gains to prevent wheel lock-up, while the H∞ component provides robust disturbance rejection against parametric uncertainties such as mass variations and sensor noise. Control efforts are distributed via a Quadratic Programming (QP) torque allocator featuring anti-windup mechanisms and explicit saturation handling to compensate for lateral drift during μ-split braking. Validation is conducted through a Model- in-the-Loop (MIL) to Software-in-the-Loop (SIL) pipeline using scenarios including wet surfaces and panic braking. Simulation results demonstrate enhanced yaw stability and controlled deceleration profiles compared to conventional baselines, ensuring computational feasibility for automotive Electronic Control Units (ECUs).
Cubillos, Ximena Celia Méndez
Ammonia is receiving heightened attention as a carbon-neutral and hydrogen energy carrier alternative fuel for compression ignition engines. However, replacing diesel with ammonia poses significant challenges due to its low reactivity and slow-burning nature, particularly at low-load conditions. This study investigated the effect of ammonia energy share (AES) on the combustion characteristics and performance of an ammonia–diesel dual-fuel (ADDF) compression ignition engine operating under low loads and at a constant speed of 1800 RPM. The experiments were conducted at three different loads: 6 Nm, 13.5 Nm, and 18 Nm, corresponding to 11%, 25%, and 33% of full load, respectively. At each load, the AES was incrementally increased, ranging from zero to its maximum limit, while maintaining the COV of IMEP below 3% to ensure stable combustion. Furthermore, CFD simulations were performed using a CONVERGE CFD model of the engine to analyze the in-cylinder thermal and chemical behavior, and the model was validated against the experimental data. The experimental results showed that the AES reached 40%, 58%, and 61% for engine loads of 6 Nm, 13.5 Nm, and 18 Nm, respectively. Increasing AES reduced the mean in-cylinder temperature and peak cylinder pressure, and shifted the peak pressure location toward the expansion stroke. Combustion phasing was delayed, and combustion duration increased with higher ammonia substitution. CFD analysis revealed weaker high-temperature and OH reaction zones, along with reduced OH and H radical activity, and increased persistence of NH2 and HO2 evolution at higher AES, indicating slower oxidation of the ammonia-containing mixture. The results highlight the challenges associated with high-ammonia operation at low loads and provide deeper insight into the combustion processes governing ADDF engine performance.
Sardar, Gobinda, Kishore, Kislay, Pradeep, P., Mittal, Mayank
Advanced Driver Assistance Systems (ADAS) are increasingly prevalent in light vehicles, both in the United States and worldwide. Moreover, ADAS are steadily being incorporated into regulatory requirements globally. Like ADAS, the automotive aftermarket is also increasing in size and significance. As both ADAS and the aftermarket industry are growing, the effect of aftermarket modifications on ADAS functionality should be examined. However, there is very little information available in the public domain about the effect of aftermarket modifications on original equipment ADAS. This work is centered on a considerable research project that was conducted to address the knowledge gap at the intersection of ADAS and the aftermarket. The project investigates five light vehicles that are important to the aftermarket, including four pickup trucks and one sport-utility vehicle. It focuses solely on the effect of popular aftermarket suspension modifications, and it does not evaluate aftermarket ADAS equipment. Typical suspension modifications were applied to the test vehicles in five modification categories, including stock, lower kits, level kits, 3–4 in. lift kits, and 6 in. lift kits. Six ADAS test procedures were performed for the test vehicles, comprised of blind spot detection, crash imminent braking, lane departure warning, pedestrian automatic emergency braking, rear cross traffic alert, and traffic jam assist. The physical tests were developed based on National Highway Traffic Safety Administration (NHTSA) New Car Assessment Program (NCAP) written experimental procedures. Statistical hypothesis testing was performed for the purpose of determining if average measured dynamic responses varied in the modified vehicles compared to stock. The results show that vehicles modified with typical aftermarket modifications will likely retain their ADAS functionality, given the limitations of the small sample size of five vehicles. Vehicles with 6 in. lift kits are expected to exhibit greater variability in their dynamic responses compared to stock. Plans for future work and unanswered research questions are outlined, with the goal of advancing aftermarket ADAS integration and ensuring the safety and performance of modified vehicles.
Bastiaan, Jennifer M., Morales, Luis, Muller, Mike
This SAE Information Report provides a broad summary of existing Reverse Automatic Emergency Braking test protocols to help assess whether additional test protocols are needed. Eventually, the task force may develop additional protocols to support testing of Reverse Automatic Emergency Braking systems.
Active Safety and Driver Support Systems Standards Committee
Internal recirculating ball screws are widely used as linear motion components in automotive active safety systems, owing to their simple structure and compact size. The recirculation (or deflection) channel is a key feature that distinguishes this type from other ball screw designs. The objective of this article is to investigate this key feature that has been rarely addressed in existing research on internal ball screw. The conventional design method for the recirculation channel involves sweeping the cross-section along the center curve. The center curve is typically defined by various classical equations. These equations are applied in different application scenarios. In automotive braking systems, high loads and strict size constraints place critical demands on both the recirculation channel and its center curve. As a representative best-practice example, the machined channel in the screw is typically employed in this application. This article compares several classical center curve equations and proposes a new general approach based on a family of transition curves. The mechanical analysis identifies the inherent structural characteristics of recirculation channel and develops corresponding design guideline. Furthermore, parameter optimization is performed using MSC ADAMS Multibody Dynamics (MBD) software.
Xia, Xinan, Xia, Yanzhe, Zhao, Tina
Ultrasonic sensors are widely deployed in automotive driver assistance systems for near-range environment perception and provide safety-relevant inputs for functions such as parking assistance and automated parking. With increasing vehicle automation, the integrity and availability of ultrasonic sensor data become more critical, as compromised measurements may lead to incorrect vehicle decisions and hazardous behavior. While prior research has extensively studied physical attacks on ultrasonic sensors, a structured cybersecurity risk analysis in accordance with automotive cybersecurity standards, combined with experimental validation, is largely missing. In particular, the communication interface between ultrasonic sensors and control units has received limited attention despite its relevance as a potential attack surface. This paper presents a systematic security analysis of an automotive ultrasonic sensing system based on a demonstrator setup. The work applies a Threat Analysis and Risk Assessment methodology aligned with ISO/SAE 21434 and HEAVENS 2.0 to identify security-relevant assets, threat scenarios, and attack paths. Risk levels are derived by evaluating potential impact and attack feasibility. To validate the risk assessment, a structured test strategy is developed using the ISTQB test process and translated into laboratory experiments. Both digital attacks targeting the sensor communication interface, with DSI3 selected as the representative protocol, and physical manipulations of the sensor environment are examined. Experimental results show that selected communication-level attacks can be realized with moderate effort and can cause controlled falsification or loss of measurement data. Physical environmental manipulations significantly degrade signal quality but do not fully suppress object detection in the evaluated configuration. The findings largely confirm the initial risk assessment while enabling refinement of attack feasibility parameters. The results provide a validated linkage between automotive cyber-security risk assessment methods and practical testing of ultrasonic sensing systems and underline the importance of jointly addressing communication interfaces and physical effects in future security concept development.
Gahm, Sebastian, Haller, Jonathan, Kriesten, Reiner
The development and validation of advanced driver-assistance systems (ADAS) and automated driving systems (ADS) are shifting from traditional linear V-model processes toward more iterative engineering cycles. Despite faster iteration, these safety-critical systems remain subject to stringent regulations. Standards and guidance, including UNECE UN Regulation No. 157 and ISO/TS 5083, emphasize traceability, transparency, and explainability throughout development and validation. Nevertheless, as ADAS/ADS are developed and validated in faster, more iterative release cycles, additional stakeholders become involved and new explainability requirements emerge. These requirements vary between stakeholders and across development, validation, and post-market deployment phases, yet they are not systematically captured in the current state of research and practice. Therefore, to ensure that explainability supports rapid iteration, it is essential to identify relevant stakeholders and specify their explainability needs. Standards such as IEEE Standard 7001-2021 provide a broad foundation for transparency in autonomous systems. However, their generic nature does not address the domain-specific complexities of ADAS/ADS. Furthermore, a conceptual gap remains between general transparency principles and explainability requirements in automotive development and validation. Building on IEEE Standard 7001-2021, this paper first offers a stakeholder taxonomy in the context of ADAS/ADS, then proposes a stakeholder-oriented analysis of explainability requirements within an automated driving use case and contexts. This analysis specifically focuses on the motivations for requiring explainability and the necessary explanation modalities. Finally, the paper discusses the limitations of the analysis and outlines directions for future research. The results of the paper provide a structured guideline for stakeholder-oriented explainability requirements in ADAS/ADS.
Liu, Xuanheng, Bairy, Akhila, Paudel, Bijay, Adolph, Laurenz, Heck, Melanie, Hettich, Lennard, Nägele, Ann-Therese, Rudolf, Korbinian, Bause, Katharina, Düser, Tobias, Schwammberger, Maike
This paper investigates the integration of Artificial Intelligence (AI) within radar-based perception for Advanced Driver Assistance Systems (ADAS) under safety considerations aligned with ISO 26262 [1] for functional safety and ISO 21448 (SOTIF) [2] for performance-related safety of the intended functionality. The study evaluates a hybrid architecture in which AI-based perception modules are combined with deterministic supervisory mechanisms to maintain safety compliance. A simulation-based case study using CARLA with radar sensor modeling is presented to compare a deterministic radar perception pipeline with an AI-enhanced approach under nominal and degraded environmental conditions. Performance is evaluated using precision, recall, and F1 score metrics. Results indicate improved recall and F1 score under adverse scenarios for the AI-based perception module, accompanied by a moderate increase in false positives. The paper discusses architectural constraints required to limit non-deterministic behavior, including confidence gating, deterministic supervision, and scenario-based validation. The findings are limited to simulation and are intended to provide preliminary insights into the technical and safety implications of incorporating AI-based radar perception within ISO 26262-compliant ADAS architectures.
Jain, Yesha
Driver monitoring systems are an important component of active safety systems, continuously evaluating the driver’s state and issuing real-time warnings. As defined by the SAE Levels of Automation, driving tasks are increasingly transferred from the driver to the vehicle from Level 0 to Level 2, however, the driver remains fully responsible for monitoring the driving environment. Current implementations, such as driver drowsiness and attention warning, assess driver alertness, while advanced driver distraction warning ensures that the driver maintains visual focus. Nevertheless, these systems do not identify the specific objects or regions the driver is observing. This limitation motivates the presented research question: can an in-car monitoring system be integrated with external environment perception sensors to infer the driver’s field of view (FoV)? This paper presents a system consisting of a driver-facing camera and a front-view camera. Facial features, including gaze direction, head pose, and iris offset are extracted using computer vision techniques. These features, together with cropped eye images, are used as inputs to a multi-modal network. Training labels were generated using a driving simulator study with 16 participants who sequentially fixated on visual targets displayed on a front screen. Experimental results show that the proposed system can predict driver visual attention and approximate FoV with a mean pixel error of 35.40 px, enabling identification of the regions of the road scene observed by the driver in real time. This work provides a foundation for explicitly modeling driver perception and its correspondence with vehicle perception systems.
Ji, Dejie, Lausch, Hendryk, Flormann, Maximilian, Henze, Roman
The current document is a part of an effort of the Active Safety Systems Committee, Active Safety Systems Sensors Task Force whose objectives are to: a Identify the functionality and performance you could expect from active safety sensors b Establish a basic understanding of how sensors work c Establish a basic understanding of how sensors can be tested d Describe an exemplar set of acceptable requirements and tests associated with each technology e Describe the key requirements/functionality for the test targets f Describe the unique characteristics of the targets or tests This document will cover items (a) and (b).
Active Safety and Driver Support Systems Standards Committee
Automotive Engineering: June 202626AUTP066/4/2026
New York 2026: diversity on full display New powertrain choices keep popping up on new vehicles from OEMs that debuted at NYIAS this year. Sealing integrity in a Formula 1 limited-slip differential High-temperature hydraulic control in a Formula 1 drivetrain requires dimensional stability, controlled sealing force, and resistance to wear under sustained pressure cycling. Inside the limited-slip differential, the sealing architecture plays a defined mechanical role in maintaining consistent torque management under race conditions. From ADAS to autonomy How engineering thermoplastics can advance sensor-based technologies. Synthetic data and the future of ADAS validation Why ADAS validation can't be solved with more miles alone. Intelligent power distribution will change the way vehicles are designed Electronic fuse (eFuse) technology can create electronic power distribution modules (ePDMs) for architectural flexibility, higher reliability, greater safety, and proactive maintenance. Editorial Maybe more than ever, let's talk transportation diversity The Navigator Can legacy automakers finally succeed with SDVs? AI scares and excites cybersecurity professionals at WCX Expert claims war hurting China's already-struggling economy NHTSA open to negotiated rulemaking on some safety issues Resilient propulsion strategies require options Driven: Honda Fastport eQuad Prototype Product Briefs Spotlight: Connectors & harnesses, EV thermal management Q&A Neural Concept's Thomas von Tschammer: Working with AI at speed
In the two months since Microvision bought Luminar and acquired key tech and talent, the sensor company has been busy. In that time, they've merged key lidar units from each company and created a perception software stack to run it in a convincing demo of its ADAS and autonomous capabilities. The company is also pushing innovative lidar tech into the defense drone and antidrone markets, already working with a German defense supplier that works with NATO member countries.
Clonts, Chris
Why ADAS validation can't be solved with more miles alone. Modern advanced driver assistance systems (ADAS) are expected to operate reliably across an almost limitless range of real-world conditions, including changing weather, low lighting, unpredictable traffic behavior, and sensor noise. Validating performance across that level of variability has become one of the most demanding parts of ADAS development. Physical road testing, or even large-scale simulation, alone cannot provide sufficient coverage to meet these demands. This challenge is driven by increasing system complexity. Modern ADAS platforms rely on machine-learning-based perception, multi-sensor fusion, and tightly integrated software architectures that must interpret complex sensor data in real time. Each additional sensing modality, software update, or feature expansion drives a significant validation effort and, in many cases, increases the number of scenarios that must be evaluated.
Kuehnke, Lutz
How engineering thermoplastics can advance sensor-based technologies. Electronics are fundamental to today's vehicles, providing critical functions from power distribution and infotainment to safety and security. Advanced driver assistance systems (ADAS) are electronic, sensor-based technologies designed to improve driver comfort and convenience by automating routine tasks and to enhance safety and reduce accidents through automated alerts, braking, and steering. Important ADAS technologies include cameras, light detection and ranging (lidar) and, especially, radio detection and ranging (radar). As the automotive industry moves toward autonomous driving, reliable signal detection and processing become even more critical. Radar provides advantages for signal detection, including capability in darkness and bad weather conditions like snow, rain and fog that can block sensors and impair visibility in optical systems. Cameras and lidar contribute better image resolution for improved identification of objects surrounding the vehicle. These technologies continue to evolve and advance.
Gisbergen, Jos van
Bird accidental collision with overhead transmission lines poses a threat to the ecology of rare bird populations. This article analyzes the warning measures to prevent birds from accidental collisions at home and abroad. In response to the low efficiency of manual installation and the poor static warning effect in preventing birds from accidental collisions with overhead transmission lines, the visual characteristics of birds are analyzed. A drone-based automatic installation flash-type bird accidental collision warning device is proposed, which includes a fixture, a disc, and a luminous circuit. The fixture can be carried and installed on the overhead line by a drone and can be easily disassembled. The disc adopts eye-catching colors and has a hollow structure to reduce wind resistance load. The luminous circuit includes solar panels, charge and discharge control circuits, flicker control circuits, batteries, and luminous components. The drone suspension warning device test was conducted, and the results showed that the device can be easily suspended from the overhead line by the drone.
Wang, Jian, Wang, Xiulong, Liu, Bin, Li, Danyu, Xu, Xunjian
Items per page:
1 – 50 of 2179