Forschungsdaten

Browse

Now showing 1 - 20 of 62

Efficient Map-Based Consistency Check Using a Sigma-Point Signed-Distance Estimator
(Universität Ulm, 2026-01-21) Wolff, Jasper; Wodtko, Thomas; Buchholz, Michael
Reliable environment perception is crucial for the functional safety and integrity of automated driving systems. However, sensor artifacts, such as reflections or partial occlusions, can generate detections that appear locally plausible but are inconsistent with the static environment. This makes digital maps a valuable prior for identifying such implausible observations. We present a probabilistic method to assess the spatial consistency between uncertain sensor detections and digital map data in automated driving. The task is formulated as the probability that an uncertain detection, transformed under an uncertain ego pose, intersects a building polygon. Since this nonlinear intersection is analytically intractable, we propose a deterministic sigma-point–based estimator that approximates the probability without random sampling. Experiments on synthetic and real urban data demonstrate that the proposed estimator achieves near Monte Carlo accuracy at real-time rates, reducing computation time by up to 45 times. The resulting probabilities enable robust identification of implausible detections in tracking and fusion pipelines.
Spectroscopic Fingerprinting of Extracellular Vesicles from Diverse Cellular Origins by ATR-FTIR: Toward Vibrational Biomarkers of Vector-Host Interactions.
(Universität Ulm, 2026) Diaz de Leon Martinez, Lorena; Groß, Rüdiger Martin
Extracellular vesicles (EVs) are nanoscale lipid bilayer structures that facilitate intercellular communication across biological systems. Although extensively studied in mammals, their spectral and biochemical characteristics in invertebrate hosts relevant to viral transmission remain poorly understood. In this study, Attenuated Total Reflectance–Fourier Transform Infrared (ATR-FTIR) spectroscopy was used to characterize and discriminate EVs derived from human dermal fibroblasts (FibEVs), hepatocytes (HepEVs), and Aedes albopictus mosquito cells (MosqEVs), alongside synthetic EV-like vesicles (SynEVs). Spectral data were analyzed using Principal Component Analysis (PCA), Canonical Analysis of Principal Coordinates (CAP), and sparse Partial Least Squares–Discriminant Analysis (sPLS-DA). PCA revealed clear clustering according to EV origin, while CAP showed strong group separation (R² ≈ 0.99), with the first two canonical axes explaining 89% of total variance. sPLS-DA identified discriminant wavenumbers associated with lipids, proteins, and nucleic acids, achieving 93% classification accuracy and an average AUC of 0.99. MosqEVs displayed lipid-enriched spectral profiles consistent with insect membrane composition, whereas mammalian EVs were protein-dominant. These findings demonstrate that ATR-FTIR is a rapid, label-free approach for EV discrimination and provide the first vibrational characterization of mosquito-derived EVs, supporting future applications in vector biology, infection surveillance, and exosome quality assessment.
Tuto3-PAT Delphi Consultation Data Anonymized
(Universität Ulm, 2026-01-07) Ramona Hiltensperger
The data set contains data of a two-round Delphi process with field experts to establish consensus on essential peer support training elements and additional features such as delivery modalities, accreditation procedures, and admission criteria. The study involved an online survey with peer support experts in several countries. The study was conducted as part of the Erasmus + funded project Tuto3-PAT.
Radar Data and Trajectories for High-Resolution Imaging of Internal Glacial Structures with a UAV-Based Radar System
(Universität Ulm, 2025-12-04) Kanz, Julian; Grathwohl, Alexander; Fischer, Michael; Bormuth, Frederik; Damm, Christian; Waldschmidt, Christian
Glacier mass loss due to rising temperatures strongly affects the Earth’s hyrdo- and ecosystem. Firn, the uppermost layer of the glacier, is the most sensitive to climatic influences and plays a crucial role for paleoclimatic research. With the UAV-based radar system presented in this work, the structure of the firn can be measured with a vertical resolution of 3–5cm. In combination with short revision times, decisive data for climate research can be provided. Measurement results from the Aletsch Glacier in Switzerland are presented, showing the ability to detect both the glacier surface and the internal structure of the firn, including the first melt layer several meters below the surface.
Matlab Files for Iterative Recovery Algorithms for Complex-Valued Distributed Compressed Sensing
(Universität Ulm, 2025-11-27) Sterk, Elena; Fischer, Robert F. H.
Distributed compressed sensing aims at the joint reconstruction of sparse signals with a common support. In some applications, complex-valued signals and sensing matrices are present. In this paper, we investigate recovery algorithms for complex-valued distributed scenarios. To that end, we review a compact exposition of complex- and vector-valued MMSE estimators. These can be used in approximate-message-passing-type algorithms. We explain joint reconstruction via iterative algorithms and evaluate suitable recovery algorithms. The performance of these algorithms is evaluated by numerical simulations for different scenarios.
SAR-Data and Trajectories for UAV-based repeat-pass interferometric SAR generation of digital elevation models (DEMs)
(Universität Ulm, 2025) Mustieles-Perez, Victor; Kim, Sumin; Kanz, Julian; Bonfert, Christina; Grathwohl, Alexander; Krieger, Gerhard; Villano, Michelangelo
Synthetic aperture radar (SAR) interferometry (InSAR) is a remote sensing technique that allows the generation of digital elevation models (DEMs) by combining two complex SAR images taken with a certain across-track separation. DEMs are of paramount importance for monitoring the Earth's surface and have traditionally been obtained using space- and airborne SAR systems, which are, however, characterized by stringent bandwidth regulations, constrained revisit times, and costly deployment and operation. Drone-borne InSAR enables cost effective surveillance of local areas with unprecedented accuracy and resolution through the exploitation of wider bandwidths and has therefore emerged as an attractive complementary system with short deployment times and flexible revisit intervals. This paper shows how to generate accurate, high-resolution DEMs through multiple acquisitions performed by a radar with ultrawide fractional bandwidth onboard a drone. We discuss the impact of the characteristics of the system and the acquisition parameters on the quality of the resulting DEMs. We also present a SAR processing scheme based on the omega-k algorithm, which allows fast processing of the drone-based raw SAR data and still yields comparable quality as back-projection algorithms, and an InSAR algorithm that takes into consideration the wide bandwidth of the system and exploits it to support phase unwrapping using radargrammetry. An experimental demonstration is conducted showing that DEMs with a height accuracy of a decimeter can be obtained at an independent posting of 25 cm × 25 cm.
Producing High-Quality Digital Elevation Models from Ultra-Wideband Drone-Borne SAR Interferometry
(Universität Ulm, 2025) Mustieles-Perez, Victor; Kim, Sumin; Kanz, Julian; Bonfert, Christina; Grathwohl, Alexander; Widmann, Marius; Bormuth, Frederik; Krieger, Gerhard; Villano, Michelangelo
Across-track synthetic aperture radar (SAR) interferometry (InSAR) is a remote sensing technique that exploits the phase difference between two complex SAR images to generate digital elevation models (DEMs), which are essential for monitoring the Earth's surface. DEMs have been obtained on a large scale from space- and airborne InSAR systems, which have, however, stringent bandwidth regulations, constrained revisit times, and costly deployment and operation. A complementary approach is addressed in this work and consists of obtaining DEMs from SAR images acquired by drone-borne radars. It enables the surveillance of local areas with unprecedented resolution and height accuracy by exploiting wider range bandwidths and larger baselines with respect to the flight height. However, due to low flight altitudes and stability issues, DEM generation presents significant challenges, especially in the case of SAR images acquired from antennas located on distinct platforms or from the same platform at different times. We propose a SAR processing scheme based on the omega-k algorithm that allows fast processing of the drone-based raw SAR data and an InSAR algorithm that takes into consideration the wide bandwidth of the system and exploits it to support phase unwrapping using radargrammetry. The system capabilities are demonstrated through experimental acquisitions over flat and hilly areas, which resulted in DEMs with a height accuracy of a decimeter at a posting of 25 cm x 25 cm.
Radar Data and Trajectories for Bistatic UAV-Based Repeater SAR for 3D Object Localization
(Universität Ulm, 2025-11-07) Kanz, Julian; Bonfert, Christina; Riekenbrauck, Ron; Waldschmidt, Christian
An unmanned aerial vehicle (UAV) based radar system is augmented by a low-complexity repeater on a second UAV that allows to locate objects in 3D space. The repeater provides complementary height information, that cannot be gathered from the monostatic radar signal alone, since linear synthetic aperture radar (SAR) fails to resolve all spatial dimensions. The high-resolution SAR image determined by the radar system can then be evaluated at the correct height, resulting in a precise object localization. Based on simulation and measurement, it is shown that the position of objects can be determined with an accuracy of a few centimeters.
Snapshot on Long COVID in Young Adults: Fast Screening Using Electronic Noses_Data
(Universität Ulm, 2025) Diaz de Leon Martinez, Lorena ; Mizaikoff, Boris
Background: Long COVID (LC) is a multisystemic condition characterized by persistent symptoms following SARS-CoV-2 infection. Although most research focuses on older or hospitalized individuals, young adults are frequently overlooked despite significant effects on their academic, professional, and social functioning. Methods: This cross-sectional study evaluated 78 university students (median age 20 years; 56.4% female) with prior COVID-19 infection, classified according to the WHO Delphi consensus definition. Sociodemographic, clinical, and spirometry data were collected, and exhaled breath samples were analyzed using an electronic nose system (e-Nose) under controlled conditions. Chemometric and machine learning techniques—Principal Component Analysis (PCA), Partial Least Squares–Discriminant Analysis (PLS-DA), Canonical Analysis of Principal Coordinates (CAP), and Random Forest (RF)—were applied to identify LC-associated volatile organic compound (VOC) patterns. Findings: LC prevalence was 29.5%. Acute-phase fatigue (OR=3.22), dyspnea (OR=6.09), nausea (OR=3.57), and vomiting (OR=11.37) were significantly associated with LC. Post-acute anosmia (OR=3.65), sleep disturbances (OR=4.34), and bradycardia were also more frequent among LC cases. All participants exhibited normal spirometry. e-Nose data effectively discriminated LC from controls (PCA variance 94.1%; CAP R²=0.95; PLS-DA accuracy 97.4%, Q²=0.534). The RF model achieved an out-of-bag error of 3.42% and ROC AUC of 0.966. Interpretation: Nearly one-third of young adults experience LC despite normal pulmonary function, suggesting substantial subclinical and systemic alterations. e-Nose breath analysis represents a promising, non-invasive, and rapid approach for LC screening, capable of supporting early detection and management in underrecognized populations.
Supplementary Material of Ph.D. Dissertation "Advancements in the Theory of Solvent Extraction and the Characterization of Annular Centrifugal Contactors"
(Universität Ulm, 2025-08-22) Hamamah, Zaid Alkhier
This material contains all the developed MATLAB software and codes, numerical values of the included graphs, and raw collected data.
SAR-Data and Trajectories for Histogram-Based Analysis of UAV-SAR Data for Agricultural Vegetation Classification
(Universität Ulm, 2025-07-25) Bormuth, Frederik; Riekenbrauck, Ron; Kanz, Julian; Sterk, Elena; Schmidt, Daniel,; Fischer, Robert F.H; Krieger, Gerhard; Waldschmidt, Christian; Damm, Christian
Climate change poses new challenges and efficiency requirements for the agricultural sector. This necessitates the coordination of fertilization and irrigation efforts by monitoring crops. Precision agriculture aims to achieve this using modern technology including unmanned aerial vehicles (UAVs). This paper shows a histogram-based approach to distinguish between crops by evaluating data of a high-performance UAV-based dual band radar at 1-4 GHz and 6-9 GHz. Three agricultural fields, were selected for this study, one each for maize, barley, and wheat. High-resolution synthetic aperture radar (SAR) images are generated for each crop in order to differentiate them based on amplitude distributions, as well as elevation dependencies.
Experimental Investigation of foaming behavior of ammonia-containing wastewater
(Universität Ulm, 2025-05-12) Lorraine Arzner
A simple experimental procedure is developed to investigate foaming of ammonia-containing wastewater in a lab-scale environment, mainly consisting of a measuring cylinder and a peristaltic pump. Experiments are performed at 600 rpm and 25 °C for 10 min, then foam decrease over time is observed. Foam is characterized by the initial foam height and the half time of the foam.
RadarLRC: A Low-Resolution Radar Dataset for Radar Clutter Detection
(Universität Ulm, 2025-04-04) Kopp, Johannes; Kellner, Dominik
The RadarLRC dataset aims to support the research on radar environment perception for autonomous driving. It is a large collection of radar point clouds measured by two automotive frequency-modulated continuous-wave (FMCW) low-resolution short range radar sensors (SRRs). The radars are mounted on a test car driving in real-world public road traffic in Germany. In addition to the radar point clouds from the SRRs, i.e. their post-CFAR detection lists, camera images showing the surroundings of the test vehicle and the vehicle's positioning data are provided. The dataset is composed of two parts. In the test set, "keyframes" of one scan per radar every 5 seconds are labeled regarding the task of radar clutter detection. This means that each individual point in the respective radar scan is assigned to one of the classes "moving object", "stationary" or "clutter". These identify points corresponding to any type of moving road user, points resulting from returns off of stationary obstacles or the ground, and erroneous points whose position and velocity do not match any of the real objects in the environment, respectively. The second part of the dataset is a supplementary collection of unlabeled samples from the same sensor setup, which can be used, for example, as a training set for a neural network approach leveraging semi-supervised or unsupervised learning. In total, the RadarLRC dataset provides recordings of about 3:52 hours of driving, resulting in more than 554k radar scans and approximately 23.33M points. Of these, 29.9k points are labeled for radar clutter detection.
The automation concept of the U-Shift II Vehicle: digital twin in Carla
(Universität Ulm, 2025-04-01) Schumann, Oliver; Wodtko, Thomas; Authaler, Dominik; Buchholz, Michael
Autonomous driving technology enables new and innovative driverless vehicle concepts to emerge, like U-Shift. Designed from the ground up, the U-Shift II platform, called driveboard, exemplifies the advantages of separating a vehicle's driving capability from the intended transportation task. It allows different so-called capsules, such as public transport or cargo, to be transported using the same U-shaped driving platform. The driveboard can change the capsules autonomously, thus providing high flexibility for fleet operators. This novel approach introduces new challenges to the task of autonomous driving. On one hand, changing sensor and vehicle configurations, e.g., when transporting a capsule with its own sensors to compensate for occlusions of the driveboard sensors by the capsule itself, requires an adaptive approach to environmental perception. On the other hand, different environments and driving tasks, as well as the augmented motion capabilities of the driveboard, require novel motion planning and control algorithms that also adapt to changing vehicle configurations. For example, the driveboard's automatic pick-up of capsules places high demands on perception and planning precision. In this paper, we first present our automation concept for the U-Shift II vehicle, which comprises environment perception and motion planning, emphasizing the modifications and enhancements compared to state-of-the-art autonomous vehicles. Second, we present a proof of concept focusing on flexibility and adaptivity based on a software-in-the-loop simulation using CARLA.
Supplementary MATLAB Code for doctoral thesis Heat Conduction in Filled Polymers - Experimental and Simulative Investigations on Microscopic Heat Transport and Particle-Level Phenomena
(Universität Ulm, 2024-11-26) Roser, Oliver
In the doctoral thesis "Heat Conduction in Filled Polymers - Experimental and Simulative Investigations on Microscopic Heat Transport and Particle-Level Phenomena", the effects of the microscopic packing structure on the effective thermal conductivity of particle-filled polymers are discussed. This dataset contains supplementary Matlab® code for microscale modeling of heat conduction in filled polymers. It contains scripts for modeling stochastic filler packings in representative volume elements (RVEs) as well as scripts for iterative calculation of steady-state heat conduction across these RVEs. Furthermore, scripts for the evaluation and determination of the effective thermal conductivity (ETC) are included. Filler packings with arbitrarily shaped particles in user-defined particle size distributions can be modeled and analyzed. The calculation results can provide valuable insights into the heat transport processes at particle level and thus contribute to a fundamental understanding of various microstructural parameters and their impact on ETC.
Scanning electrochemical probe microscopy: towards the characterization of micro- and nanostructured photocatalytic materials
(Universität Ulm, 2024-06-24) Giada, Caniglia; Sarah, Horn; Christine, Kranz
Platinum-black (Pt-B) has been demonstrated as an excellent electrocatalytic material for the electrochemical oxidation of hydrogen peroxide (H2O2). As Pt-B films can be deposited electrochemically, micro- and nano-sized conductive transducers can be modified with Pt-B. Here, we present the potential of Pt-B micro- and sub-micro-sized sensors for the detection and quantification of hydrogen (H2) in solution. Using these microsensors, no sampling step for H2 determination is required and e.g., in photocatalysis, the onset of H2 evolution can be monitored in situ. We present Pt-B- based H2 micro- and sub-micro-sized sensors based on different electrochemical transducers such as microelectrodes and atomic force microscopy (AFM)- scanning electrochemical microscopy (SECM) probes, which enable local measurements e.g., at heterogenized photocatalytically active samples. The microsensors are characterized in terms of limits of detection (LOD), which ranges from 4.0 µM to 30 µM depending on the size of the sensors and the experimental conditions such as type of electrolyte and pH. The sensors were tested for the in situ H2 evolution by light-driven water-splitting, i.e., using ascorbic acid or triethanolamine, showing a wide linear concertation range, good reproducibility, and high sensitivity. Proof-of-principle experiments using Pt-B-modified cantilever-based sensors were performed using a model sample like platinum substrate to map the electrochemical H2 evolution along with the topography using AFM-SECM.
Replication Package for: Textual Queries for Graphical Roadmap Models with Edit History
(Universität Ulm, 2024-04-22) Jutz, Benedikt
The appended ZIP file contains the Replication Package to the master's thesis "Textual Queries for Graphical Roadmap Models with Edit History". Its contents are (listing adapted from the thesis, Appendix A): • The folder "LimeSurvey Questionnaires" contains the created LimeSurvey questionnaire with the scenarios (Subsection 7.3.2) and other forms for SUS, TLX and demographic data (Section 7.5), This questionnaire is provided as print version (Printed Survey.pdf), and as runnable version (Executable Survey.lss). • NewBusCoach.json contains the bus coach model. • The folder "R Scripts" contains six analysis files that can be run using the R programming language: calculateTOTScenario.R, calculateTCR.R, calculateSUS.R, calcRTLX.R, showDemograhics.R, and calculateATI.R. They provide all qualitative and demographic data summaries and diagrams in Chapter 8. We also include the exported results from LimeSurvey, before inferring TCR values (results-survey374125_copy.csv), and afterwards in the file results-survey374125_copy_after_tcr_analysis.csv. • The folder "User Guide" contains the user guide (Subsection 7.3.3), both as readable PDF (User_Guide_EditQL.pdf), and in raw form (User_Guide_EditQL.tex). The "Images" subfolder contains all images in the guide. • The "Videos" folder contains our two sets of videos: tutorial videos to be watched before the study (subfolder "Tutorials"), and scenario videos to be watched during the study (subfolder "Scenarios"). Screen recordings are available at request. • Our study protocols, and the TeX template, are in the folder "Study Protocols".
Research data to the paper "Screening instruments of cognition: the relation of the Mini-Mental State Examination to the Edinburgh Cognitive and Behavioural ALS Screen in amyotrophic lateral sclerosis"
(Universität Ulm, 2024-04-18) Finsel, Julia; Lulé, Dorothée
This is the raw research data for Serian et al. Screening instruments of cognition: the relation of the Mini-Mental State Examination to the Edinburgh Cognitive and Behavioural ALS Screen in amyotrophic lateral sclerosis. It provides all cognitive and disorder-related data regarding the included patients of the study.
Forschungsdaten zu "Exploratory Study on how Programmers interact with LLM based Chatbots"
(Universität Ulm, 2023-12-22) Haas, Yves
These is the data for a qualitative think-aloud study on how programmers interact with LMM-based chatbots. It contains synchronized transcripts of the think-aloud recordings and the recordings themselves without audio. It contains the codes they are coded with as well. The data is in form of a MaxQDA20 (.mx20) project and is part of a master thesis with the title "Exploratory Study on how Programmers interact with LLM based Chatbots". The transcribed recordings are only identified by a random number which can not be linked to the study participants.
Using convolutional neural networks for stereological characterization of 3D hetero-aggregates based on synthetic STEM data: FAIR data
(Universität Ulm, 2023-11-07) Fuchs, Lukas
A dataset of Synthetic STEM-image data of various hetero-aggregates is presented. The aggregates consist of TiO2 and WO3 primary particles. The aggregates are generates using a stochastic 3D model which is based on cluster-cluster aggregation. A physic based simulator is used to generate the STEM-images.

Browse

Recent Submissions