TY - JOUR A1 - Moser, Elisabeth A1 - Jobst, Simon A1 - Bierl, Rudolf A1 - Jenko, Frank T1 - A Deep Learning System to Transform Cross-Section Spectra to Varying Environmental Conditions JF - Vibrational Spectroscopy N2 - Absorption cross-sections provide a basis for many gas sensing applications. Therefore, any error in molecular cross-sections caused by varying environmental conditions propagates to spectroscopic applications. Original molecular cross-sections in varying environmental conditions can only be simulated for some molecules, whereas for most multi-atom molecules, one must rely on high-precision measurements at certain environmental configurations. In this study, a deep learning system trained with simulated absorption cross-sections for predicting cross-sections at a different pressure configuration is presented. The system’s capability to transfer to measured, multi-atom cross-sections is demonstrated. Thus, it provides an alternative to (pseudo-) line lists whenever the required information for simulation is unavailable. The predictive performance of the system was evaluated on validation data via simulation, and its transfer learning capabilities were demonstrated on actual measurement chlorine nitrate data. From the comparison between the system and line lists, the system shows slightly worse performance than pseudo-line lists but its predictive quality is still deemed acceptable with less than 5% relative integral change with a highly localized error around the peak center. This opens a promising way for further research to use deep learning to simulate the effect of varying environmental conditions on absorption cross-sections. KW - Spectroscopy KW - Deep learning KW - Simulation KW - Pressure broadening KW - Pseudo-line lists KW - Transfer learning Y1 - 2022 U6 - https://doi.org/10.1016/j.vibspec.2022.103410 SN - 0924-2031 VL - 122 IS - September PB - Elsevier ER - TY - CHAP A1 - Moser, Elisabeth A1 - Pangerl, Jonas A1 - Jobst, Simon A1 - Weigl, Stefan A1 - Bierl, Rudolf T1 - Modeling the Photoacoustic Spectrum of a Quantum Cascade Laser for Human Breath T2 - Optical Sensors and Sensing Congress 2022 (AIS, LACSEA, Sensors, ES): July 11 - 15, 2022, Vancouver, British Columbia Canada N2 - A modeling approach to create a photoacoustic spectrum from synthetic data is presented and evaluated. The resulting model reaches a MAPE score of 2.7% and can be used to enable data-driven development in future work. KW - Quantenkaskadenlaser KW - Photoakustische Spektrometrie KW - Atemluft Y1 - 2022 SN - 978-1-957171-10-4 U6 - https://doi.org/10.1364/AIS.2022.ATu3G.2 PB - Optica Publishing Group ER - TY - JOUR A1 - Goldschmidt, Jens A1 - Moser, Elisabeth A1 - Nitzsche, Leonard A1 - Bierl, Rudolf A1 - Wöllenstein, Jürgen T1 - Improving the performance of artificial neural networks trained on synthetic data in gas spectroscopy – a study on two sensing approaches JF - tm - Technisches Messen N2 - Artificial neural networks (ANNs) are used in quantitative infrared gas spectroscopy to predict concentrations on multi-component absorption spectra. Training of ANNs requires vast amounts of labelled training data which may be elaborate and time consuming to obtain. Additional data can be gained by the utilization of synthetically generated spectra, but at the cost of systematic deviations to measured data. Here, we present two approaches to train ANNs with a combination of comparatively small, measured data sets and synthetically generated data. For the first approach a neural network is trained hybridly with synthetically generated infrared absorption spectra of mixtures of N2O and CO and measured zero-gas spectra, taken with a mid-infrared dual comb spectrometer. This improves the mean absolute error (MAE) of the network predictions from 0.46 to 0.01 ppmV and 0.24 to 0.01 ppmV for the concentration predictions of N2O and CO respectively for zero-gas measurements which was previously observed for training with purely synthetic data. At the same time a similar performance on spectra from gas mixtures of 0–100 ppmV N2O and 0 to 60 ppmV CO was achieved. For the second approach an ANN pre-trained on synthetic infrared spectra of mixtures of acetone and ethanol is retrained on a small dataset consisting of 26 spectra taken with a mid-infrared photoacoustic spectrometer. In this case the MAE for the concentration predictions of ethanol and acetone are improved by 45 % and 20 % in comparison to purely synthetic training. This shows the capability of using synthetically generated data to train ANNs in combination with small amounts of measured data to further improve neural networks for gas sensing and the transferability between different sensing approaches. N2 - Künstliche neuronale Netze werden in der quantitativen Infrarot-Gasspektroskopie eingesetzt, um Gaskonzentrationen in Multi-Komponenten-Absorptionsspektren vorherzusagen. Das Training von künstlichen neuronalen Netzen benötigt große Mengen an Trainingsdaten, was mit einem hohen Arbeits- und Zeitaufwand verbunden ist. Zusätzliche Trainingsdaten können durch synthetisch generierte Spektren gewonnen werden, sind jedoch mit systematischen Abweichungen zu tatsächlich gemessenen Daten verbunden. In dieser Arbeit werden zwei Ansätze präsentiert, um künstliche neuronale Netze mit einer Kombination aus vergleichsweise kleinen, gemessenen Datensätzen und synthetisch erzeugten Daten zu trainieren. Für den ersten Ansatz wird das neuronale Netz hybrid mit synthetisch erzeugten Infrarot-Absorptionsspektren von Mischungen aus N2O und CO und gemessenen Null-Gas-Spektren trainiert, welche mit einem Doppel-Kamm-Spektrometer im Mittelinfraroten aufgenommen wurden. Dieses Vorgehen verbessert den mittleren, absoluten Fehler der Netzvorhersagen von 0,46 auf 0,01 ppmV und 0,24 auf 0,01 ppmV für N2O bzw. für CO für Null-Gas-Messungen, im Vergleich zu den Vorhersagen des neuronalen Netzes, welches nur mit synthetisch erzeugten Daten trainiert wurde. Gleichzeitig zeigte das hybrid trainierte Netz eine gleichbleibende Performance für Konzentrationsvorhersagen für Spektren aus Gasmischungen von 0 bis 100 ppmV N2O und 0 bis 60 ppmV CO. Für den zweiten Ansatz wird ein neuronales Netz, welches bereits mit synthetischen erzeugten Infrarotspektren von Mischungen aus Aceton und Ethanol trainiert wurde, erneut auf einem kleinen gemessenen Datensatz bestehend aus 26 Spektren trainiert, welche mit einem photoakustischen Spektrometer im mittleren Infrarot aufgenommen wurden. In diesem Fall hat sich der mittlere absolute Fehler der Konzentrationsvorhersage für Ethanol und Aceton um 45 % bzw. 20 % im Vergleich zu einem rein synthetischen Training verbessert. Das zeigt das Potential synthetisch erzeugte Daten für das Training von künstlichen neuronalen Netzen in Kombination mit kleinen gemessenen Datensätzen zu nutzen, um neuronale Netze für den Einsatz im Bereich der Gassensorik zu verbessern und gleichzeitig die Übertragbarkeit der Methodik auf verschiedene Messansätze. Y1 - 2023 U6 - https://doi.org/10.1515/teme-2023-0051 PB - Oldenbourg Wissenschaftsverlag ET - Online veröffentlicht ER - TY - CHAP A1 - Ulreich, Fabian A1 - Moser, Elisabeth A1 - Olbrich, Florian A1 - Ebert, Martin A1 - Bierl, Rudolf A1 - Kaup, André ED - Keil, Rudolf ED - Tschorn, Jan Alexander ED - Tümler, Johannes ED - Altinsoy, Mehmet Ercan T1 - Luminance Simulation in CARLA under Cloud Coverage - Model Validation and Implications T2 - 2023 IEEE International Workshop on Metrology for Automotive (MetroAutomotive), -30 June 2023, Modena, Italy N2 - To decrease the number of kilometers driven during the development of autonomous cars or driving assistance systems, performant simulation tools are necessary. Currently, domain distance effects between simulation and reality are limiting the successful application of rendering engines in data-driven perception tasks. In order to mitigate those domain distance effects, simulation tools have to be as close to reality as possible for the given task. For optical sensors like cameras, the luminance of the scene is essential. We provide within this paper a method to measure the luminance of rendered scenes within CARLA, an often used open-source simulation environment. Thereby, it is possible to validate the environment and weather models by taking real-world measurements with photometric sensors or with the help of open-source weather data, published e.g. by the German federal service for weather data (DWD - "Deutscher Wetterdienst"). Employing our proposed luminance measurement, the domain gap resulting from the simulation can be specified, which makes it possible to evaluate the statements about the safety of the automated driving system determined within the simulation. We show that the ratio between global and diffuse radiation modeled by the default atmosphere models within CARLA are under limited conditions similar to real-world measurements taken by the DWD. Nevertheless, we show, that the ratio’s temporal variability in real-world situations is not modeled by CARLA. KW - Atmospheric measurements KW - Atmospheric modeling KW - Cameras KW - carla KW - Cloud computing KW - cloud coverage KW - Data models KW - luminance KW - meteorology KW - Simulation KW - Soft sensors KW - weather KW - Wind Y1 - 2023 SN - 979-8-3503-2187-6 U6 - https://doi.org/10.1109/MetroAutomotive57488.2023.10219098 SP - 228 EP - 233 PB - IEEE ER - TY - JOUR A1 - Müller, Max A1 - Weigl, Stefan A1 - Müller-Williams, Jennifer A1 - Lindauer, Matthias A1 - Rück, Thomas A1 - Jobst, Simon A1 - Bierl, Rudolf A1 - Matysik, Frank-Michael T1 - Comparison of photoacoustic spectroscopy and cavity ring-down spectroscopy for ambient methane monitoring at Hohenpeißenberg JF - Atmospheric Measurement Techniques N2 - With an atmospheric concentration of approximately 2000 parts per billion (ppbV, 10−9), methane (CH4) is the second most abundant greenhouse gas (GHG) in the atmosphere after carbon dioxide (CO2). The task of long-term and spatially resolved GHG monitoring to verify whether climate policy actions are effective is becoming more crucial as climate change progresses. In this paper we report the CH4 concentration readings of our photoacoustic (PA) sensor over a 5 d period at Hohenpeißenberg, Germany. As a reference device, a calibrated cavity ring-down spectrometer, Picarro G2301, from the meteorological observatory of the German Weather Service (DWD) was employed. Trace gas measurements with photoacoustic instruments promise to provide low detection limits at comparably low costs. However, PA devices are often susceptible to cross-sensitivities related to fluctuating environmental conditions, e.g. ambient humidity. The obtained results show that for PA sensor systems non-radiative relaxation effects induced by varying humidity are a non-negligible factor. Applying algorithm compensation techniques, which are capable of calculating the influence of non-radiative relaxation effects on the photoacoustic signal, increase the accuracy of the photoacoustic sensor significantly. With an average relative deviation of 1.11 % from the G2301, the photoacoustic sensor shows good agreement with the reference instrument. Y1 - 2023 U6 - https://doi.org/10.5194/amt-16-4263-2023 SN - 1867-8548 VL - 16 IS - 18 SP - 4263 EP - 4270 PB - Copernicus Publications ER - TY - INPR A1 - Pangerl, Jonas A1 - Moser, Elisabeth A1 - Müller, Max A1 - Weigl, Stefan A1 - Jobst, Simon A1 - Rück, Thomas A1 - Bierl, Rudolf A1 - Matysik, Frank-Michael T1 - A Highly Sensitive Acetone and Ethanol Quantum Cascade Laser Based Photoacoustic Sensor: Characterization and Multi-Component Spectra Recording in Synthetic Breath T2 - SSRN Electronic Journal N2 - Trace gas analysis in breath is challenging due to the vast number of different components. We present a highly sensitive quantum cascade laser based photoacoustic setup for breath analysis. Scanning the range between 8260 and 8270 nm with a spectral resolution of 48 pm, we are able to quantify acetone and ethanol within a typical breath matrix containing water and CO2. We photoacoustically acquired spectra within this region of mid-infra-red light and prove that those spectra do not suffer from non-spectral interferences. The purely additive behavior of a breath sample spectrum was verified by comparing it with the independently acquired single component spectra using Pearson and Spearman correlation coefficients. A previously presented simulation approach is improved and an error attribution study is presented. With a 3σ detection limit of 6.5 ppbV in terms of ethanol and 250 pptV regarding acetone, our system is among the best performing presented so far. KW - Photoacoustic spectroscopy KW - quantum cascade laser KW - spectral simulation KW - breath analysis KW - acetone Y1 - 2022 U6 - https://doi.org/10.2139/ssrn.4305376 PB - Elsevier ER - TY - PAT A1 - Olbrich, Florian A1 - Pongratz, Christian A1 - Bierl, Rudolf A1 - Ehrlich, Ingo T1 - Method and System for Evaluating a Structural Integrity of an Aerial Vehicle Y1 - 2024 ER - TY - CHAP A1 - Peller, Sebastian A1 - Zankl, Tobias A1 - Fischer, Christoph A1 - Bierl, Rudolf T1 - Fast sound field characterization of beamforming capable capacitive micromachined ultrasonic transducer (CMUT) arrays by refracto-vibrometry T2 - IEEE IUS 2023, International Ultrasonics Symposium, Montréal, September 3-8, 2023 N2 - We introduce a time-optimized setup based on refracto-vibrometry for the purpose of scanning the sound field of capacitive micromachined ultrasonic transducer (CMUT) arrays primarily for the qualification of their beamforming capability. In comparison to commonly used microphones on a traversing stage the proposed method is substantially faster and consumes only a few minutes of time for a complete two-dimensional sound field containing about 100.000 scan points. T2 - Schnelle Schallfeld-Charakterisierung von CMUT-Arrays mithilfe von Refrakto-Vibrometrie KW - CMUT KW - LDV KW - refracto-vibrometry KW - sound field KW - beamforming Y1 - 2023 SN - 979-8-3503-4645-9 U6 - https://doi.org/10.1109/IUS51837.2023.10308126 SN - 1948-5727 SP - 1 EP - 3 PB - IEEE ER - TY - JOUR A1 - Rück, Thomas A1 - Bierl, Rudolf A1 - Lechner, Alfred A1 - Graf, Antonia A1 - Dams, Florian A1 - Schreiner, Rupert A1 - Auchter, Eberhard A1 - Kriz, Willy A1 - Deubzer, MIchael A1 - Schiller, Frank A1 - Mottok, Jürgen A1 - Niemetz, Michael A1 - Margull, Ulrich A1 - Hagel, Georg A1 - Utesch, Matthias A1 - Waldherr, Franz A1 - Böhm, Matthias A1 - Fraunhoffer, Judith A1 - Gardeia, Armin A1 - Schneider, Ralph A1 - Streubel, Janet A1 - Landes, Dieter A1 - Studt, Reimer A1 - Peuker, Dominik A1 - Scharfenberg, Georg A1 - Hook, Christian A1 - Schuster, Dietwald A1 - Ehrlich, Ingo A1 - Dinnebier, Heinrich A1 - Briem, Ulrich A1 - Lämmlein, Stephan A1 - Koder, Alexander A1 - Bialek, Adam A1 - Genewsky, Axel A1 - Neumeier, Michael A1 - Schlosser, Philipp A1 - Rabl, Hans-Peter A1 - Paule, Matthias A1 - Galster, Christoph A1 - Schiedermeier, Michael A1 - Zwickel, Andreas A1 - Hobmeier, Christoph A1 - Bischoff, Tobias A1 - Rill, Georg A1 - Schaeffer, Thomas A1 - Arbesmeier, Martin A1 - Groß, Andreas A1 - Schlegl, Thomas A1 - Becker, Mark A1 - Senn, Konrad A1 - Schliekmann, Claus A1 - Scholz, Peter A1 - Sippl, Christian A1 - Grill, Martin ED - Eckstein, Josef T1 - Forschungsbericht 2011 / Hochschule für Angewandte Wissenschaften - Fachhochschule Regensburg T3 - Forschungsberichte der OTH Regensburg - 2011 Y1 - 2011 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:898-opus4-7321 SN - 1868-3533 CY - Regensburg ER - TY - CHAP A1 - Roth, Carina A1 - Hausler, Peter A1 - Bierl, Rudolf T1 - Einbindung eines 2D-Photodetektors in einen hochintegrierten SPR-Imaging-Sensor T2 - 7. MikroSystemTechnik Kongress "MEMS, Mikroelektronik, Systeme", 23.-25. Oktober 2017, München N2 - Die Oberflächenplasmonenresonanzspektroskopie (SPR) ist eine hochempfindliche Messmethode, die es erlaubt, Gase und Flüssigkeiten zerstörungs- und markierungsfrei in Echtzeit zu analysieren. Bisher vornehmlich im Labormaßstab in der Bioanalytik und dem Wirkstoffscreening eingesetzt, soll diese Technologie nun miniaturisiert und für weitere Anwendungsgebiete zugänglich gemacht werden. Dazu wird ein kompakter Micro-Opto-Electro-Mechanical Systems Sensor (MOEMS) entwickelt, der mit Hilfe des SPR Imaging Änderungen der chemischen Zusammensetzung verschiedener Flüssigkeiten inline messen kann. N2 - Surface plasmon resonance spectroscopy (SPR) is a highly sensitive measurement method which allows the analysis of gases and liquids in real-time in a non-destructive and marking-free manner. Presently SPR is predominantly used for pharmaceutical screening and biotechnical analysis. Now this technology is to be miniaturized and made accessible for further application areas. For this purpose, a compact micro-opto-electro-mechanical system sensor (MOEMS) is developed, which can measure changes in the chemical composition of different fluids using SPR imaging. T2 - Integration of a 2D-photodetector in a SPR-Imaging-Sensor KW - Bildsensor KW - Bioanalytik KW - chemische Zusammensetzung KW - Fotodetektor KW - Labormaßstab KW - Messmethode KW - mikrooptoelektromechanisches System KW - Oberflächenplasmon KW - Oberflächenplasmonenresonanzspektroskopie KW - pharmazeutisches Screening KW - Spektroskopie KW - Wirkstoff Y1 - 2017 SN - 978-3-8007-4491-6 SP - 848 EP - 850 PB - VDE-Verlag CY - München ER - TY - CHAP A1 - Bauer, Lukas A1 - Vitzthumecker, Thomas A1 - Bierl, Rudolf A1 - Ehrnsperger, Matthias T1 - Machine-learning-based detection and severity estimation of drought stress in plants using hyperspectral imaging data T2 - Remote Sensing for Agriculture, Ecosystems, and Hydrology XXVII N2 - Growing food demand due to population growth, coupled with increasingly frequent and severe droughts caused by climate change make water increasingly scarce. To address this, accurate assessment of plant water demand is essential for precise drought treatment and water conservation. Hyperspectral imaging (HSI) captures hypercubes, a combination of spectral and spatial data and offers promising capabilities for detection of plant stresses. However, most reported approaches only use selected spectral bands or indices, neglecting the full hypercube information. This is assumed to limit the detection accuracy. To overcome these limitations, we aim to develop a measurement pipeline to generate a comprehensive dataset comprising hypercubes of plants under varying drought stress levels along with selected physiological, environmental, and illumination data. This dataset will be used to train suitable data-driven models that enable improved drought stress detection as well as the non-invasive determination of physiological parameters based on HSI data. Y1 - 2025 U6 - https://doi.org/10.1117/12.3072011 PB - SPIE ER - TY - INPR A1 - Fischer, Johannes A1 - Hirsch, Thomas A1 - Reitmeier, Torsten A1 - Bierl, Rudolf T1 - Real-time hardware-based processing of high-precision detector signals for surface plasmon resonance spectroscopy N2 - Surface plasmon resonance (SPR) is limited by small-signal detectability and drift when subtraction occurs in software after digitization. We introduce an SPR detector that performs on-detector amplification and analog differential readout, eliminating moving parts and software-heavy correction. The hardware-native subtraction boosts the usable ADC range and suppresses illumination and environmental noise. In fixed-angle refractive-index steps (NaCl), the platform resolves Δn_min ≈ 1.8 × 10⁻⁷ RIU compared to 4.6–7.2 × 10⁻⁶ RIU on a commercial comparator and improves small-signal SNR by up to ∼5,000-fold, while remaining competitive at high signal levels. In a model IgG–BSA assay, the detector’s low noise floor clarifies early binding and equilibrium transitions. By generating inherently clean raw signals, this hardware-native approach dramatically enhances sensitivity and long-term stability for label-free biosensing and inline process analytics while rendering AI-based or complex post-processing entirely unnecessary. The concept generalizes across platforms and opens a compact route to robust, high-fidelity SPR in complex environments, with a clear path toward multi-wavelength and arrayed detectors for high-throughput chemical monitoring. KW - Drift-free SPR KW - Hardware-native sensing KW - Noise suppression KW - On-detector amplification KW - Small-signal detection KW - Surface plasmon resonance Y1 - 2025 U6 - https://doi.org/10.2139/ssrn.5971170 ER -