FSP2: Mobilität
Refine
Year of publication
Document Type
- conference proceeding (article) (170)
- Article (124)
- conference proceeding (presentation) (28)
- Preprint (13)
- Part of a Book (10)
- Master's Thesis (8)
- Report (7)
- conference proceeding (volume) (3)
- Book (2)
- conference proceeding (summary, abstract) (2)
Language
- English (312)
- German (58)
- Multiple languages (2)
Publication reviewed
- peer review (187)
- begutachtet (113)
- nicht begutachtet (70)
Keywords
- automated driving (14)
- advanced driver assistance systems (10)
- Simulation (9)
- human factor (8)
- workload (8)
- driving experience (7)
- Automated Driving (6)
- FEB NAFV (6)
- lane keeping system (6)
- stress (6)
Institute
- IFM - Institut für Fahrerassistenz und vernetzte Mobilität (144)
- Fakultät Elektrotechnik (133)
- Fakultät Maschinenbau (88)
- Fakultät Tourismus-Management (52)
- INIT – Institut für Nachhaltige und Innovative Tourismusentwicklung (52)
- Fakultät Informatik (51)
- IFA – Institut für Innovative Fahrzeugantriebe (28)
- IMS - Institut für Maschinelles Sehen (20)
- ILE – Institut für Leistungselektronik (10)
- EPT – Institut für Effiziente Produktionstechnik (8)
Explainable Statistical Evaluation of Emergency Braking Functions in Scenario-Based Safety Testing
(2026)
This article presents the integration of the Explainable Statistical Evaluation method (ESE) with an Autonomous Emergency Braking (AEB) using the software library FASim in the context of the European New Car Assessment Programme Car-to-Pedestrian Nearside Child Obstructed 50% scenario (Euro NCAP CPNCO-50 scenario). ESE integrates qualitative and quantitative safety methods for an easy to explain justification of functional safety (FuSa) and safety of the intended functionality (Sotif) for complex systems. The article addresses the European Union’s call for objective, coherent safety metrics for Connected and Automated Vehicles and shows how simulation-based test generation can support a statistically sound argument for safety, for which validation tests do not provide sufficient evidence.
PEMFC durability is still a major challenge. To overcome time-consuming durability tests, so-called accelerated durability tests (ADT) are of urgent need. This work presents our recent results in developing ADT protocols in the context of realistic operating conditions, i.e. voltage clipping at 0.85 V. A 5,500 h long-term test was carried out as reference applying a realistic automotive drive cycle. Focusing on different stressors such as temperature, relative humidity (RH) and load profile four different ADT protocols of 1,200 h duration were derived. 7-cell short stacks with 240 cm² active area were used. Comparing cell voltage as key indicator, an acceleration factor of 3 to 7 could be achieved. In situ characterization techniques such as spatially resolved current measurement, CV and EIS were employed to investigate influences of individual stressors on specific degradation mechanisms and components. Highest acceleration was observed in mass transport region of ADTs addressing RH as stressor, suggesting that RH cycling leads to increased degradation of hydrophobic surfaces. Increased temperature was found to accelerate primarily carbon support degradation. Accelerated catalyst aging seems to be low, demonstrating the effectiveness of voltage clipping conditions. Our most promising ADT shows quite homogeneous acceleration of voltage degradation across all current regions.
Synthetic Data Generation for AI-Based Quality Inspection of Laser Welds in Lithium-Ion Batteries
(2025)
Manufacturing companies are increasingly confronted with critical challenges such as a shortage of skilled labor, rising production costs, and ever-stricter quality requirements. These challenges become particularly acute when defect types exhibit high visual variance, making consistent and accurate inspection difficult. Traditionally, visual inspection of high variance errors is performed manually by human operators—a process that is both costly and prone to errors. Consequently, there is a growing interest in replacing human inspection with AI-based visual quality control systems. However, the adoption of such systems is often hindered by limited access to training data, labor-intensive labeling processes, or the absence of real production data during early development stages. To address these challenges, this paper presents a methodology for training AI models using synthetically generated image data. The synthetic images are created using Physically Based Rendering, which enables precise control over rendering parameters and facilitates automated labeling. This approach allows for a systematic analysis of parameter importance and bypasses the need for large real training datasets. As a case study, the focus is on the inspection of laser welds in battery connectors for fully electric vehicles—a particularly demanding application due to the criticality of each weld. The results demonstrates the effectiveness of synthetic data in training robust AI models, thereby providing a scalable and efficient alternative to traditional data acquisition and labeling methods. The trained binary classifier reaches a precision of 0.94 with a recall of 0.98 solely trained on synthetic data and tested on real image data.
A common goal of unpaired image-to-image translation is to preserve content consistency between source images and translated images while mimicking the style of the target domain. Due to biases between the datasets of both domains, many methods suffer from inconsistencies caused by the translation process. Most approaches introduced to mitigate these inconsistencies do not constrain the discriminator, leading to an even more ill-posed training setup. Moreover, none of these approaches is designed for larger crop sizes. In this work, we show that masking the inputs of a global discriminator for both domains with a content-based mask is sufficient to reduce content inconsistencies significantly. However, this strategy leads to artifacts that can be traced back to the masking process. To reduce these artifacts, we introduce a local discriminator that operates on pairs of small crops selected with a similarity sampling strategy. Furthermore, we apply this sampling strategy to sample global input crops from the source and target dataset. In addition, we propose feature-attentive denormalization to selectively incorporate content-based statistics into the generator stream. In our experiments, we show that our method achieves state-of-the-art performance in photorealistic sim-to-real translation and weather translation and also performs well in day-to-night translation. Additionally, we propose the cKVD metric, which builds on the sKVD metric and enables the examination of translation quality at the class or category level.
The contamination of vehicle camera lenses by artifacts such as mud, dust, or rain droplets presents a safety critical challenge that must be addressed by Advanced Driver Assistance Systems (ADAS). These obstructions, which impair the camera's functionality, are categorized and analyzed into three distinct classes: opaque, translucent, and water droplets. Using the publicly available Soiling-WoodScape (1) dataset, we investigate the biases and effects associated with these different types of contaminations. Our analysis focuses on the position, shape, and other attributes of the affected regions. This detailed understanding enables the development of improved simulation methods and provides valuable insights into the biases inherent in learning based models trained on this dataset. Such knowledge is essential for enhancing the robustness and reliability of ADAS under real world soiling conditions.
This article investigates path planning strategies for autonomous vehicles in critical pedestrian scenarios, using a digital presentation of a real scenario based on the ASAM Open Simulation Interface® (OSI) standard. We present a comparative study of two decision-making algorithms -an Occupancy Grid Method (OG) and an Artificial Potential Field Method (APF) -applied to the Euro NCAP CPNCO-50 scenario, a critical use case for autonomous emergency braking systems. Simulations are implemented using OSI to enable modular and standardized integration across simulation platforms.
The OG Method reacts preemptively to potential collisions by detecting obstacles within a discretized environment model, initiating early evasive maneuvers and offering conservative, safety-oriented responses. In contrast, the APF Method adapts dynamically by modeling repulsive risk potentials, resulting in behavior more similar to that of human drivers.
The framework allows parameter tuning to reflect different driving styles and can incorporate Predictive Potential Field Method (PPF) that anticipate future trajectories. This enables efficient algorithm comparison and iteration. Real-world scenarios can be resimulated using OSI trace file to validate virtual performance against physical tests.
The rapid growth of the global elderly population presents significant challenges for ensuring safe and inclusive mobility. Personal Mobility Vehicles offer critical support for independent travel but often lack interfaces adapted to age-related declines in vision, motor control, and cognition. This study evaluates ten PMV models, selected based on regulatory compliance in Germany and Japan, using a structured scoring framework grounded in human-machine interface design principles for older adults. The results reveal significant usability deficits, particularly in lever ergonomics, display readability, and control intuitiveness, that can increase cognitive load and contribute to operational errors. These human errors, in turn, pose serious safety risks in public and transit environments. The findings underscore the need for standardized, user-centered interface design to reduce the likelihood of human error and enhance overall operational safety. By addressing these deficiencies, PMVs can better support aging users and contribute to the broader goal of achieving zero traffic accidents.
This project presents an investigation into the effect of camera choice on single camera trilateration (SCT) using computer vision for ego vehicle localisation. It focuses on the performance of the SCT algorithm, using three cameras—LUCID ATL089S-CC, IDS U3-3280CP-MGL, and APTINA (Onsemi) AR0132AT—and utilises a state-of-the-art computer vision pipeline for semantic segmentation and monocular depth estimation (MDE).
Classic computer vision algorithms are employed to detect keypoints, and ground truth correspondences are established using the iNovitas infra3D Web-Client. The analysis reveals that camera choice has a minimal effect on SCT performance, although factors such as calibration and resolution influence the accuracy of depth estimation and keypoint selection.
Perceived reliability, safety and comfort benefits are key factors of customer satisfaction with advanced driver assistance systems. Satisfaction again is linked to technology trust, acceptance and diffusion. Therefore, it plays a pivotal role for advancing road safety and for the future market penetration for advanced driver assistance systems of level 3 and beyond. By understanding the interplay between customer expectations, satisfaction, and behaviour, manufacturers can refine systems to address not just convenience but also critical safety concerns. Empirical results underscore that improving customer satisfaction directly contributes to higher trust, usage, and thus increased road safety, aligning with FAST-zero's mission of achieving zero accidents.
Customer satisfaction results from a subjective comparison of expectations and experiences. With the help of the disconfirmation paradigm (Confirmation/Disconfirmation paradigm), customer expectations can be identified and characterized. The matching process is subjective, as cognitive and affective factors influence the resulting satisfaction -the customer draws on existing experience, information and knowledge. Affectiveemotional driven factors in the context of driver assistance systems are for example, the perceived feeling of safety or the enjoyment of driving.
This conference paper is based on an extensive quantitative survey (sample: 609 participants in Germany), which was supplemented with qualitative interviews. It presents the survey results on automatic distance control and the lane keeping assistant.
Durability of proton exchange membrane fuel cells (PEMFCs) in single-cell setup is often assessed using potentiostatic test protocols under inert cathode atmosphere, such as the US DOE’s accelerated stress tests (ASTs). Although these voltage-controlled protocols impose consistent electrochemical stress, they do not accurately replicate real-world operating conditions, for instance due to the lack of product water formation in the cathode or due to the generation of heat. In contrast, stress testing under hydrogen and air, which is commonly employed for PEMFC durability testing on stack level, reflect the hybrid potential- and current-controlled degradation modes which are typical of mobile applications. This is usually done galvanostatic, but can also be realized in a voltage controlled manner. There are notable differences in the ageing behaviors between those two methods: Potentiostatic testing protocols, in which the voltage remains constant throughout the test duration, induce consistent electrochemical stress regarding potential driven degradation modes such as platinum dissolution, Ostwald ripening, carbon corrosion and chemical membrane degradation due to hydrogen peroxide formation. In contrast, galvanostatic testing protocols result in a gradual decline in potential with time, thereby reducing potentially driven electrochemical stress. However, in galvanostatic mode, water production remains constant (unlike in potentiostatic mode), which is known to contribute equally to fuel cell degradation. This dynamic of varying potential driven and humidity driven stress significantly affects catalyst and membrane durability, underscoring the need to understand the degradation behavior specific to each testing method. The present study compares the impact of load-cycle and voltage-cycle protocols on PEMFC degradation. Electrochemical characterizations, such as cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS), are used to identify the ageing behavior of the individual components. All tests were performed using commercial Gore® PRIMEA® membrane electrode assemblies (MEA) in a Baltic quickConnect high amp test cell with an active area of 12 cm². Our findings highlight the importance of testing under hydrogen and air to better replicate operational conditions and enhance PEMFC durability in practical applications.