Repository der Technischen Hochschule Ingolstadt
Not a member yet
4170 research outputs found
Sort by
Does mental model similarity equal innovation team performance?
No full textPurpose
This paper addresses challenges and opportunities of interdisciplinary teamwork in innovation management, focusing on how team mental models (TMMs) can foster more effective collaboration among team members from diverse backgrounds. The research examines the relationship between TMM similarity and team performance in interdisciplinary innovation teams.
Design/methodology/approach
An exploratory study was conducted with 55 participants across 15 teams. TMM similarity was measured by analyzing team interactions using a web-based interface that facilitated contrasted comparison ratings. These ratings were first processed into individual graphs via a shortest-node-finding pathfinder algorithm, then compared. Subject matter experts evaluated team performance.
Findings
The results reveal a significant positive correlation between TMM similarity and team performance in interdisciplinary innovation teams, suggesting that greater alignment in team members' mental models enhances overall innovation project performance.
Research limitations/implications
Although the academic programs sampled replicate the environment, challenges and various other aspects of innovation projects, they can only be considered proxies for innovation projects within real organizations. Further research within professional environments, using a larger sample, is recommended.
Practical implications
The findings highlight the value of assessing and fostering TMM similarity to improve teamwork and performance in interdisciplinary innovation projects. The interface and code used are publicly available to encourage their implementation in organizations.
Originality/value
This research provides novel insights into the application of TMMs within interdisciplinary innovation teams, extending the concept beyond its traditional use in unidisciplinary and structured task settings
YoFlow Method for Scenario Based Automatic Accident Detection
No full textRecent advances in sensor and computing technologies have enabled road side units (RSUs) to not only monitor traffic flow but also process data in real time to improve road safety. However, leveraging RSUs for proactive accident detection remains a challenging and underexplored task, partly due to the lack of diverse accident data. To address this, this study proposes two key contributions: (i) a scenario-based synthetic data generation framework, and (ii) YoFlow, a novel system for vehicle-tovehicle accident detection from a simulated RSU camera perspective. The proposed framework leverages the PEGASUS method for scenario generation strategy and BeamNG.tech for generating synthetic traffic videos. This approach led to the development of the SB-SIF dataset, which includes five representative intersection crash scenarios derived from German accident data. The SB-SIF dataset contains 914 crash videos, 123 near-miss events, and 924 normal traffic instances and is publicly available at: https://doi.org/10.5281/zenodo.15267252. The proposed YoFlow system identifies accidents by analyzing temporal variations in vehicle speed vectors, using YOLO for vehicle classification and CUDA-accelerated dense optical flow to capture abrupt motion changes. The extracted features are processed and classified using an XGBoost model, achieving 94% recall and 90% precision in accident detection
An Environmental Management Information System for Closing Knowledge Gaps in Corporate Sustainable Decision-Making
No full textKooperative intelligente Verkehrssysteme für die forensische Unfallanalyse
No full textDie vorliegende Dissertation befasst sich mit den Chancen und Herausforderungen für die forensische Unfallanalyse, die sich durch den Einsatz kooperativer intelligenter Verkehrssysteme (C-ITSs) ergeben. Im Zentrum der Untersuchung steht die kooperative Car-to-X-Direktkommunikation (C2X-Kommunikation) im 5,9 GHz Frequenzbereich, eine Technologie, die seit 2019 mit der Markteinführung des Volkswagen Golf 8 erstmals für den europäischen Massenmarkt verfügbar ist. Die übergeordnete Forschungsfrage der Arbeit besteht darin, wie die Chancen kooperativer intelligenter Verkehrssysteme für die forensische Unfallanalyse genutzt und gleichzeitig die entstehenden Herausforderungen adressiert werden können. Durch die erstmalige Verknüpfung dieser beiden Forschungsfelder sowie einer systematischen Analyse entlang der Schichten des Protokollstapels konnten Chancen und Herausforderungen der C2X-Kommunikation für die Unfallanalyse identifiziert werden. Darauf aufbauend sind vier Forschungsfragen formuliert worden, die in dieser Dissertation untersucht und beantwortet wurden.
Die Chancen für die Unfallanalyse liegen vor allem im Erhalt zusätzlicher digitaler Spuren durch die Fahrzeugzu-Allem-Kommunikation (C2X)-Kommunikationsdaten. Um diese Chance zu nutzen, ist es erforderlich, den Mehrwert der Daten für die Unfallanalyse nachzuweisen. Im Rahmen dieser Arbeit wurden Experimente und Analysen durchgeführt, um den Mehrwert von C2X-Kommunikationsdaten hinsichtlich Verfügbarkeit, Genauigkeit und Anwendbarkeit für die forensische Unfallanalyse zu bewerten. Besondere Aufmerksamkeit wurde der Cooperative Awareness Message (CAM) gewidmet, die wesentliche Informationen wie Fahrzeugposition, Geschwindigkeit und Fahrtrichtung enthält.
Fahrversuche wurden durchgeführt, um die Genauigkeit und zeitliche Verfügbarkeit der CAM-Daten zu erforschen und deren gerichtssichere Verwendbarkeit in der Unfallanalyse sicherzustellen. Die Ergebnisse verdeutlichen, dass CAM-Daten eine vergleichbare oder höhere Genauigkeit und höhere zeitliche Auflösung im Unfallhergang aufweisen können als die Daten des Event Data Recorders (EDR). Darüber hinaus konnte durch die Entwicklung und Anwendung eines Kalman-Filters die mediane Positionsgenauigkeit der CAM-Daten erhöht werden. Ergänzend wurde ein Rekonstruktionsprogramm entwickelt, das ausschließlich auf CAM-Daten basiert und die direkte Anwendbarkeit der CAM-Daten für die Unfallanalyse plausibilisiert. Zuletzt wurde das Konzept einer forensischen Roadside-Unit (F-RSU) vorgestellt, die eine unabhängige Speicherung und Bereitstellung von C2X-Kommunikationsdaten ermöglicht.
Neben den Chancen wurden zentrale Herausforderungen identifiziert. Dazu zählen die Berücksichtigung von C2X-Warnungen auf das Fahrerverhalten, die Integration der Daten in bestehende Rekonstruktionssoftware sowie datenschutzrechtliche Aspekte. Zur Adressierung dieser sollten primär bestehende Rekonstruktionsprogramme um C2X-Funktionen erweitert und neue forensische Grunddaten ermittelt werden. Für Letzteres wurde in dieser Arbeit eine Probandenstudie in einem Fahrsimulator durchgeführt. Die Ergebnisse zeigen, dass Fahrer auf C2X-Warnungen durch Gaswegnahme oder Bremsen angemessen reagieren, auch wenn der Warnungsgrund nicht unmittelbar erkennbar ist. Negative Auswirkungen auf die Reaktionsdauer der Probanden in Verbindung mit C2X-Warnungen konnten ebenfalls nicht nachgewiesen werden. Eine falsch-positive C2X-Warnung führte in der Studie lediglich bei 2 von 32 Probanden zu verkehrsgefährdenden Reaktionen, was darauf hindeutet, dass ihr Einfluss auf die Unfallmitursächlichkeit als gering einzustufen ist.
Die Ergebnisse der Dissertation verdeutlichen, dass C2X-Kommunikationsdaten eine entscheidende Erweiterung der digitalen Datenbasis für die Unfallanalyse darstellen können. Sie liefern nicht nur bedeutende Informationen im Unfallhergang über die Unfallbeteiligten, sondern ermöglichen dazu die Analyse des Einflusses unbeteiligter Verkehrsteilnehmer und der Verkehrsinfrastruktur.
Der wesentliche Beitrag dieser Dissertation liegt in der Schaffung einer Grundlage für die zukünftige, gerichtssichere Nutzung von C2X-Kommunikationsdaten in der Unfallanalyse. Die vollständige Ausschöpfung des Potenzials der C2X-Technologie erfordert jedoch weitere Forschungsarbeiten sowie rechtliche Klärungen hinsichtlich der Speicher- und Verarbeitungsmöglichkeiten zu forensischen Zwecken.This dissertation explores the opportunities and challenges for forensic accident analysis arising from the deployment of Cooperative Intelligent Transport Systems (C-ITSs). The investigation focuses on cooperative Vehicle-to-X short-range communication (hereafter referred to as V2X communication) in the 5.9 GHz band, a technology that has been available on the European mass market for the first time since 2019 with the introduction of the Volkswagen Golf 8. The overarching research question of this work is how the opportunities provided by C-ITSs can be utilized for forensic accident analysis while simultaneously addressing the associated challenges. By linking these two research fields for the first time and conducting a systematic analysis across the protocol stack layers, this study identifies both the opportunities and challenges of V2X communication for accident analysis. Based on these findings, four research questions were formulated, examined, and answered in this dissertation.
The opportunities for accident analysis arise primarily from the availability of additional digital traces provided by V2X communication data. To fully leverage this opportunity, it is essential to demonstrate the benefits of these data for accident analysis. In this thesis, experiments and analyses were conducted to evaluate the added value of V2X communication data in terms of availability, accuracy, and applicability for forensic accident analysis. Special emphasis was placed on the Cooperative Awareness Message (CAM), which contains essential information such as vehicle position, speed and direction of travel.
Driving tests were conducted to investigate CAM data’s accuracy and temporal availability and ensure its forensic usability in accident analysis. The results indicate that Cooperative Awareness Message (CAM) data provide a comparable or higher level of accuracy and temporal resolution for accident reconstruction than the data from the Event Data Recorder (EDR). Furthermore, the development and application of a Kalman filter enhanced the median positional accuracy of CAM data. A reconstruction program that exclusively utilizes CAM data was also developed, validating its direct applicability for accident analysis. Moreover, the concept of a forensic roadside unit (F-RSU) was introduced, enabling the independent storage and provision of V2X communication data.
In addition to the opportunities, key challenges were identified. These include the influence of V2X warnings on driver behavior, the integration of V2X communication data into existing reconstruction software, and data protection concerns. To address these challenges, existing reconstruction programs should primarily be extended to include V2X functionality and new fundamental forensic data should be determined. For the latter, a driving simulator study was conducted with 32 participants. The results indicate that drivers respond appropriately to V2X warnings by either releasing the accelerator or braking, even when the reason for the warning is not immediately apparent. Moreover, no negative effects could be observed on participants’ reaction times in response to V2X warnings. A false-positive V2X warning led to traffic-endangering reactions in only 2 out of 32 participants, indicating a minimal impact on accident causation.
The findings of this dissertation demonstrate that V2X communication data can serve as a significant extension of the digital database for accident analysis. They provide valuable information about the accident sequence involving the directly affected parties while also enabling the evaluation of the impact of uninvolved road users and surrounding traffic infrastructure. This dissertation’s main contribution is creating a basis for the future, court-proof utilization of V2X communication data in accident analysis. However, fully realizing the potential of V2X technology requires further research and legal clarification regarding storage and processing options for forensic purposes
Evaluation of a Deep Learning and XAI based Facial Phenotyping Tool for Genetic Syndromes: A Clinical User Study
No full textArtificial intelligence (AI) tools are increasingly employed in clinical genetics to assist in diagnosing genetic conditions by assessing photographs of patients. For medical uses of AI, explainable AI (XAI) methods offer a promising approach by providing interpretable outputs, such as saliency maps and region relevance visualizations. XAI has been discussed as important for regulatory purposes and to enable clinicians to better understand how AI tools work in practice. However, the real-world effects of XAI on clinician performance, confidence, and trust remain underexplored. This study involved a web-based user experiment with 31 medical geneticists to assess the impact of AI-only diagnostic assistance compared to XAI-supported diagnostics. Participants were randomly assigned to either group and completed diagnostic tasks with 18 facial images of individuals with known genetic syndromes and unaffected individuals, before and after experiencing the AI outputs. The results show that both AI-only and XAI approaches improved diagnostic accuracy and clinician confidence. The effects varied according to the accuracy of AI predictions and the clarity of syndromic features (sample difficulty). While AI support was viewed positively, users approached XAI with skepticism. Interestingly, we found a positive correlation between diagnostic improvement and XAI intervention. Although XAI support did not significantly enhance overall performance relative to AI alone, it prompted users to critically evaluate images with false predictions and influenced their confidence levels. These findings highlight the complexities of trust, perceived usefulness, and interpretability in AI-assisted diagnostics, with important implications for developing and implementing clinical decision-support tools in facial phenotyping for rare genetic diseases
Beats vs. Talks: Alleviating Virtual Reality Sickness with Music and Podcasts
No full textAs virtual reality (VR) technology advances, its potential for widespread application increases. However, VR sickness (VRS) remains a major obstacle to broader adoption. This study investigates the effect of auditory stimuli on VRS during VR experiences. It contrasts the effects of music and podcasts, with a no-sound environment serving as reference. Twenty-three participants experienced three different auditory conditions during a VR roller coaster ride in a randomized, balanced order. VRS was quantified using the Simulator Sickness Questionnaire (SSQ) and physiological data. The SSQ results indicate that music significantly reduced oculomotor disturbances and disorientation, while podcasts had no positive effects. The physiological data demonstrated no significant effects. The majority of participants preferred the music scenario, describing it as relaxing and pleasant. This highlights the potential of music, especially when self-selected and perceived as pleasant, to improve VR experience by significantly reducing VRS. This effect appears to be independent of physiological data
Communicating Uncertainty in Arrival Time Predictions for Public Transport: A Comparison of Point and Interval Forecasts
No full textIn public transport, arrival times are typically communicated as point forecasts, aiming to present precise estimates. However, current prediction models are unable to provide such precise and reliable estimates due to unpredictable events. This results in arrival times on passenger information systems appearing inaccurate due to the lack of communicated uncertainty. We therefore investigated interval forecasts as an alternative in an online study, aiming to better communicate uncertainty in arrival times. Our findings indicate that interval forecasts improve the communication of uncertainty. Further, user satisfaction was driven primarily by waiting time, and this relationship was moderated by the forecast concept. Point forecasts were only well received when the bus arrived as predicted, otherwise users preferred the broader interval forecasts. Participants valued accuracy over precision when judging arrival times
BIPV in India: Opportunities, challenges, and pathways for urban planning and smart cities
No full textAs India urbanizes rapidly, sustainable energy solutions have become a priority to meet rising energy demands and mitigate environmental impacts. Building-Integrated Photovoltaics (BIPV) offer a promising approach to harness solar energy within urban infrastructure, transforming buildings into renewable energy assets. Unlike conventional rooftop PV systems, BIPV maximizes urban space utilization while enhancing architectural aesthetics and energy efficiency. However, the adoption of BIPV in India remains limited due to factors such as regulatory barriers, lack of standardized designs, and high initial costs. This study uses SWOT (Strengths, Weaknesses, Opportunities, Threats) and TOWS (Threats, Opportunities, Weaknesses, Strengths) analyses to assess the potential, opportunities, and challenges of BIPV in India’s urban planning. It then discusses policy implications and offers practical recommendations for implementation in Indian cities. Key findings of this study indicate that BIPV adoption in India can significantly contribute to urban sustainability by reducing carbon emissions, improving energy self-sufficiency, and lowering long-term operational costs. The SWOT and TOWS analysis reveal that while BIPV presents opportunities for smart urban integration, challenges such as high initial investments and lack of awareness must be addressed through targeted policies and incentives. Additionally, global case studies highlight successful BIPV implementations, providing valuable lessons for India’s urban planning strategies
Deep Segmentation of 3+1D Radar Point Cloud for Real-Time Roadside Traffic User Detection
No full textSmart cities rely on intelligent infrastructure to enhance road safety, optimize traffic flow, and enable vehicle-to-infrastructure (V2I) communication. A key component of such infrastructure is an efficient and real-time perception system that accurately detects diverse traffic participants. Among various sensing modalities, automotive radar is one of the best choices due to its robust performance in adverse weather and low-light conditions. However, due to low spatial resolution, traditional clustering-based approaches for radar object detection often struggle with vulnerable road user detection and nearby object separation. Hence, this paper proposes a deep learning-based 3+1D radar point cloud clustering methodology tailored for smart infrastructure-based perception applications. This approach first performs semantic segmentation of the radar point cloud, followed by instance segmentation to generate well-formed clusters with class labels using a deep neural network. It also detects single-point objects that conventional methods often miss. The described approach is developed and experimented using a smart infrastructure-based sensor setup and it performs segmentation of the point cloud in real-time. Experimental results demonstrate 95.35% F1-macro score for semantic segmentation and 91.03% mean average precision (mAP) at an intersection over union (IoU) threshold of 0.5 for instance segmentation. Further, the complete pipeline operates at 43.61 frames per second with a memory requirement of less than 0.7 MB on the edge device (Nvidia Jetson AGX Orin)