We present an advanced membraneless optical microphone (MeoM)-based photoacoustic spectroscopy (PAS) system for trace gas detection. The MeoM-PAS combines Michelson interferometry with photoacoustic resonance to achieve all-optical detection by sensing minute refractive index variations between the two interferometer arms induced by photoacoustic pressure fluctuations. The all-optical design enhances stability and reduces both acoustic and electromagnetic interference. This configuration also allows the optical components to remain physically isolated from the target gas, enabling non-contact measurement. A resonant gas cell was designed to match the interferometric configuration, and an experimental platform was established to evaluate the MeoMPAS performance with NO2. Experimental results showed that the system achieved a minimum detection limit of 4.1 ppb, further improved to 2.5 ppb at an averaging time of 107 s, as determined by Allan deviation analysis. These results demonstrate the feasibility of the MeoM-PAS approach for trace gas monitoring in atmospheric and industrial applications.

We present a comprehensive optical, thermal, and mechanical characterization of two commercially available thin-clad germanosilicate single-mode fibers named "SM1500(7.8/80)P" and "SM1500(4.2/50)P" whose smaller cladding diameters are advantageous for fiber-based force sensing in comparison to a standard SMF28e+ fiber. The effective refractive indices were determined using a precise fiber Bragg grating (FBG) inscription-based method with precision in the order of 10(-5). Moreover, a numerical approach reveals GeO2 doping levels of 3.42(5) mol% for SMF28e+, 5.92(8) mol% for SM1500(7.8/80)P, and 20.09(28) mol% for SM1500(4.2/50)P, which were validated via energy-dispersive X-ray spectroscopy. The increased GeO2 content in the thin-clad fibers correlates with their enhanced photosensitivity. The temperature and strain sensitivities of the fibers were quantified, showing only minor changes in temperature response and nearly constant strain coefficients across all samples. As expected, the force sensitivity increased significantly with decreasing cladding diameter, achieving a 2.4-fold and 6.1-fold enhancement for the 80 & micro;m and 50 & micro;m fibers, respectively, compared to SMF28e+. These enhancements, achieved without post-processing such as tapering or etching, preserve the mechanical and optical integrity of the fibers while improving their force-to-temperature sensitivity ratio. The findings demonstrate that thin-clad germanosilicate fibers are promising candidates for high-performance, temperature-compensated force sensing platforms compatible with standard telecom equipment and ultra-violet FBG fabrication technologies.

Hydrogen monitoring requires accurate, thermally stable sensors for detecting low hydrogen concentrations under variable environmental conditions. We demonstrate a multi-parametric hydrogen and temperature sensor based on dual-FBG Fabry–Pérot etalons integrated into thin-clad optical fibers and functionalized with palladium–gold nanofilms. Two sensitivity-enhancement strategies were combined: reducing the fiber cross-section to improve strain transfer between nanocoating and fiber, and alloying palladium with gold to increase hydrogen solubility and reduce the miscibility gap. The optimized Pd85:Au15 coating (86 nm) on a 50μm fiber achieved 24pm/%(H2) at 20°C. We introduce the normalized sensitivity, accounting for the coating-to-fiber cross-sectional ratio, to compare the intercalation-mediated strain sensing effect across platforms in a thorough review. The sensor enabled ISO-26142-compliant simultaneous read-out from 10–60°C and 100–20,000 ppm in nitrogen. In air, the lowest measurable concentration was 3000 ppm, with transferable calibration above this threshold and excellent repeatability over 260 h of cyclic hydrogen loading.

Frequency-modulated continuous wave (FMCW) radio detection and ranging (RADAR) sensors have become indispensable technologies for automated driving systems (ADS) due to their reliability in adverse weather conditions and their ability to simultaneously measure the distance to objects, relative radial velocity, and azimuth and elevation angles. The automotive industry has increasingly considered simulation-based testing of autonomous vehicles due to safety, cost, and time constraints. This raises the need for virtual environmental perception sensors that provide results close to reality. This work presents the design and structure of a ray-tracing-based, high-fidelity, tool-independent baseband FMCW RADAR sensor model. The RADAR sensor model is developed using the standardized functional mock-up interface (FMI) and open simulation interface (OSI) and is integrated into the co-simulation environment of commercial software to demonstrate its exchangeability. The RADAR FMU model incorporates a multiple input and multiple output (MIMO) 2D linear spacing virtual antenna array, non-coherent integration (NCI) of range-Doppler maps (RDMs) over receiver antennas, a constant false alarm rate (CFAR) to obtain an interim object detection list, and density-based spatial clustering of applications with noise (DBSCAN) to provide a single detection per object. The presented RADAR FMU model also includes RADAR sensor-specific impairments such as phase noise (PN), radio frequency (RF) group delay, phase imbalance (PI) of transmitter antennas, mixer non-linearity including third-order intermodulation products (IM3), and noise figure (NF) of receiver antennas. Additionally, this work presents a methodology for plausibly verifying the RADAR sensor model at the raw data level (range map (RM) and RDM) and object detection list level. The simulation results are compared with real sensor measurements to validate the modeling of sensor-specific impairments. The mean absolute percentage error (MAPE) metric is used to quantify the difference between the simulation and real sensor measurements. The results demonstrate that the complete signal processing toolchain and sensor-specific impairments of the RADAR sensor must be considered to achieve simulation results that closely resemble those of the real sensor.

Intraspectral referencing comprises a single-sensor solution, where simultaneous measurements of hydrogen concentration and temperature are achieved within a single, partially palladium-coated pi -phase-shifted fiber Bragg grating (FBG). Highly localized, hydrogen-induced strain sections provoke the notch and the envelope of the pi -FBG spectrum to yield different hydrogen responses, while temperature affects both features almost identically. The precise temperature compensation in sub-Kelvin ranges was used to characterize the functional layer in the range of 20(degrees)C- 50(degrees)C and at 250-20 000 ppm H-2 in N-2. Upon initial hydrogen exposure after fabrication, a gradually decreasing sensitivity drift was observed, which relaxed and stabilized after sufficiently long initialization with hydrogen. The desorption consisted of at least two processes, a fast desorption when large amounts of hydrogen are present in the nanofilm, followed by a notably slow desorption (similar to 5 h) until the baseline was recovered. A zero-point referenced calibration scheme of the PTFE-capped Pd-91:Ni-09 fiber optic hydrogen sensor revealed a strong nonlinear response, especially at low concentrations <1000 ppm. This required linearization, which led to effective decoupling of the temperature cross-sensitivity through iterative data processing. Decoupled concentrations and temperatures show a lower detection limit of 250 ppm, a significant enhancement in baseline stability, and time constants of t(90) = 19 s for ab- and t(10) = 18 s for desorption. This work paves the way for both practical applications of fiber optic H-2 sensors and fundamental research in Pd nanofilm engineering.

Light detection and ranging (LiDAR) sensor technology for people detection offers a significant advantage in data protection. However, to design these systems cost- and energy-efficiently, the relationship between the measurement data and final object detection output with deep neural networks (DNNs) has to be elaborated. Therefore, this paper presents augmentation methods to analyze the influence of the distance, resolution, noise, and shading parameters of a LiDAR sensor in real point clouds for people detection. Furthermore, their influence on object detection using DNNs was investigated. A significant reduction in the quality requirements for the point clouds was possible for the measurement setup with only minor degradation on the object list level. The DNNs PointVoxel-Region-based Convolutional Neural Network (PV-RCNN) and Sparsely Embedded Convolutional Detection (SECOND) both only show a reduction in object detection of less than 5% with a reduced resolution of up to 32 factors, for an increase in distance of 4 factors, and with a Gaussian noise up to μ=0 and σ=0.07. In addition, both networks require an unshaded height of approx. 0.5m from a detected person’s head downwards to ensure good people detection performance without special training for these cases. The results obtained, such as shadowing information, are transferred to a software program to determine the minimum number of sensors and their orientation based on the mounting height of the sensor, the sensor parameters, and the ground area under consideration, both for detection at the point cloud level and object detection level.

Intelligent characterization of van der Waals semiconductors is an essential process for industrial manufacturing and laboratory fabrication. A combination of microscopic images and artificial intelligence models is an efficient way for wafer‐scale layer number identification of van der Waals semiconductors. This methodology overcomes the bottleneck of the conventional manual layer number counting approach, which requires a long period of manual inspection and induces high error rates when distinguishing layers with similar appearance. Here, a convolutional architecture that involves a fused network of ResNet‐Inception with Attention Layer (RIAL) is developed for accurate multiclass classification of randomly distributed layers of chemical vapor deposition (CVD)‐grown van der Waals semiconductors. RIAL model is first validated on the single‐label datasets CIFAR‐10/100, and subsequently fine‐tuned on the custom‐built microscopic image datasets of CVD‐grown MoS2. To compare with semantic segmentation, U‐Net with Attention Layer (UNAL) is further implemented for pixel‐wise classification of multiclass semiconductors. The quantitative analysis of RIAL and UNAL illustrates the versatility of attention convolutional network models in the wafer‐scale identification of van der Waals semiconductors.

For well-magnified imaging systems that satisfy the Nyquist criterion, camera pixels resolve the fine details provided by the objective lens. However, the mismatch in a space-bandwidth product (SBP) between the objective lenses and digital cameras leads to major sacrifices in the optical field of view (FOV). This Letter presents a framework of phase manipulating Fresnel lenses (PMFL) to bridge this SBP gap in quantitative phase imaging (QPI) modality. This framework breaks through the conventional design paradigm of interferometric QPI, converting it from the combination of individual optical elements to a direct phase transform of the optical field. The wide-field QPI is implemented by a single dynamic PMFL mask, which conducts differential interference and scans the optical FOV across the camera sensor. We built a microscopic platform equipped with the PMFL module to image both natural and artificial samples and expanded the FOV for a given camera by factors of 3 × 3, 2 × 2, and 4 × 1, respectively.

Hyperspectral imaging generates vast amounts of data containing spatial and spectral information. Dimensionality reduction methods can reduce data size while preserving essential spectral features and are grouped into feature extraction or band selection methods. This study demonstrates the efficiency of the standard deviation as a band selection approach combined with a straightforward convolutional neural network for classifying organ tissues with high spectral similarity. To evaluate the classification performance, the method was applied to eleven groups of different organ samples, consisting of 100 datasets per group. Using the standard deviation is an effective method for dimensionality reduction while maintaining the characteristic spectral features and effectively decreasing data size by up to 97.3%, achieving a classification accuracy of 97.21% compared to 99.30% without any processing. Even in comparison with mutual information- and Shannon entropy-based band selection methods, the standard deviation exhibited superior stability and efficiency while maintaining equally high classification accuracy. The results highlight the potential of dimensionality reduction for hyperspectral imaging classification tasks that require large datasets and fast processing speed without sacrificing accuracy.

The high-temperature performance of a sapphire fiber Bragg grating (FBG) has been characterized by a commercial single-mode interrogator. By using the mode field matching fusion splicing method, we proposed before, high-order modes in the reflection spectrum of the sapphire FBG could be suppressed and a quasi-single-mode reflection spectrum could be obtained from a single-mode fiber-based demodulation system. A 48-h isothermal annealing at 900 degrees C and two temperature calibration processes, one up to 800 degrees C and another up to 1200 degrees C, were carried out successively. There was no obvious reflectivity decay or wavelength drift during the 900 degrees C annealing process, and no hysteresis during the two temperature calibration processes. The precisions of the sapphire FBG were +/- 0.25 degrees C in the 800 degrees C calibration process, while increased to +/- 5 degrees C in the 1200 degrees C calibration process.

Fiber Bragg gratings (FBGs) have been extensively used for single-point and multi-point measurements, mostly inscribed in single-mode fibers. However, it is feasible to inscribe FBGs in multimode fibers, which resist bending and can perform discriminative sensing of multiple physical parameters. When using a simple experimental setup to measure the temperature dependence of the dip in the transmission spectrum, significant fluctuations in its spectral power can be observed. Therefore, this study shows that the temperature-dependent spectral power fluctuations in multimode FBGs can be mitigated using a reflectometric configuration with suppressed modal interference, leading to higher-reliability temperature sensing.

Multimode fiber (MMF) imaging aided by machine learning holds promise for numerous applications, including medical endoscopy. A key challenge for this technology is the sensitivity of modal transmission characteristics to environmental perturbations. Here, we show experimentally that an MMF imaging scheme based on a neural network (NN) can achieve results that are significantly robust to thermal perturbations. For example, natural images are successfully reconstructed as the MMF's temperature is varied by up to 50^∘C relative to the training scenario, despite substantial variations in the speckle patterns caused by thermal changes. A dense NN with a single hidden layer is found to outperform a convolutional NN suitable for standard computer vision tasks. In addition, we demonstrate that NN parameters can be used to understand the MMF properties by reconstructing the approximate transmission matrices, and we show that the image reconstruction accuracy is directly related to the temperature dependence of the MMF's transmission characteristics.

The use of LiDAR sensor technology for people detection offers a significant advantage in terms of data pro-tection. In LiDAR point clouds, unlike camera images, people can be detected but not identified without further information. LiDAR sensors are, therefore, particularly suitable for detecting people in publicly accessible places and reacting accordingly to the number of people, for example, with on demand services at airports. Due to the anonymity of people in LiDAR point clouds, personal data is protected, and approval for implementing such a detection system is simpler than that of comparable camera systems. In this paper, we present a measurement setup that covers the configuration of the sensor setup, the creation of a dataset for training neural networks for object detection, and the object detection itself. The measurement setup generates an average of 2408 automatically labeled point clouds per sensor, per hour. The SECOND network trained with this dataset achieves average precision for the intersection over union of the 2D view with a threshold of 0.5 of 87.67 %, the PV-RCNN of 85.74 % and an average precision for the average orientation similarity with a threshold of 0.5 of 89.56 %, and for the PV-RCNN of 87.81 %.

The detection of gas concentrations in harsh environments with high performance is an emerging research area. Ensuring that external factors do not affect the detection and that it is immune to electromagnetic interference are key factors in guaranteeing reliable gas detection in a variety of tough environments. Among all types of gas-sensing methods, photoacoustic (PA) methods based on optical interferometers offer high sensitivity, anti-electromagnetic interference capability, and non-magnetic saturation. This article first presents the basic principles and characteristics of the main optical interferometers, the PA effect, and the core technologies of PA gas-sensing methods. It then reviews advanced optical interferometer-based methods for PA gas sensing and summarizes their characteristics, including the heterodyne-based Mach-Zehnder interferometer (MZI), optical fiber-based MZI, Sagnac interferometer (SI)-based, diaphragm-based extrinsic Fabry-Perot (F-P) interferometer (EFPI), cantilever-based EFPI, cantilever-based Michelson interferometer (MI), difference-based MI, and membrane-less optical microphone-photoacoustic spectroscopy (MeoM-PAS) sensing methods. In particular, the MeoM-PAS methods for gas sensing are discussed in detail, as they offer a completely static measurement system and the separation of a PA gas cell from the measuring system for applications in complicated and adverse circumstances. This review also outlines the advantages and limitations of these PA gas-sensing technologies.

LiDAR sensors are crucial sensors for highly automated vehicles. Relevant information about the vehicle's surroundings can be obtained from 3D point clouds, which serves as a basis for decision-making for further control of the vehicle. For this purpose, information about the type and position of objects in the vehicle's environment and their velocity and movement direction is essential. In this paper, we present VelObPoints, a neural network architecture that estimates the longitudinal and lateral velocity and performs object detection on a single LiDAR point cloud. Compared to existing tracking methods, the neural network allows this information to be extracted from a single LiDAR frame. We propose a simulated dataset for training and testing, containing motion distortion effects. The neural network achieves a mean Intersection over Union of 0.863 and a mean average velocity error of 0.332ms(-1). Based on a single point cloud, this information, consisting of the object's scaling, position, and rotation, as well as its velocities in the longitudinal and lateral directions, is immediately available for the driving function. For subsequent motion prediction, and object tracking, leading to more quickly available velocity and motion direction information and higher redundancy in sensor data fusion.

Automated vehicles use light detection and ranging (LiDAR) sensors for environmental scanning. However, the relative motion between the scanning LiDAR sensor and objects leads to a distortion of the point cloud. This phenomenon is known as the motion distortion effect, significantly degrading the sensor’s object detection capabilities and generating false negative or false positive errors. In this work, we have introduced ray tracing-based deterministic and analytical approaches to model the motion distortion effect on the scanning LiDAR sensor’s performance for simulation-based testing. In addition, we have performed dynamic test drives at a proving ground to compare real LiDAR data with the motion distortion effect simulation data. The real-world scenarios, the environmental conditions, the digital twin of the scenery, and the object of interest (OOI) are replicated in the virtual environment of commercial software to obtain the synthetic LiDAR data. The real and the virtual test drives are compared frame by frame to validate the motion distortion effect modeling. The mean absolute percentage error (MAPE), the occupied cell ratio (OCR), and the Barons cross-correlation coefficient (BCC) are used to quantify the correlation between the virtual and the real LiDAR point cloud data. The results show that the deterministic approach matches the real measurements better than the analytical approach for the scenarios in which the yaw rate of the ego vehicle changes rapidly.

Ultra-thin atomic crystals are promising for fabricating next-generation photonic and optoelectronic devices. Wafer-scale characterization techniques are highly desired for efficient and accurate thickness identification of these crystals and their heterostructures. Optical contrast between atomic crystals and substrates based on Fresnel theory is a key technique for the identification of thicknesses. Both RGB color information and spectroscopic information have been explored for layer number counting and implemented with machine learning algorithms based on large amounts of data for feature extraction. In this work, a multispectral microscopic method combining the hardware design and deep-learning algorithms is developed. Multispectral image restoration during large-area scanning caused by optical imaging modality drifts and automated layer number identification based on multispectral grayscale images are studied using deep learning models: generative adversarial network (GAN) and 3D U-Net. These models are trained using custom-built multispectral data sets and evaluated quantitatively with indicators (Dice coefficient, confusion matrix, structural similarity). After these models are trained and tested, they are integrated into a graphic user interface for on-site identification use. The developed method provides a framework using multispectral images for 3D data reconstruction and segmentation and can be implemented for wafer-scale characterization of heterostructures containing different species of ultra-thin atomic crystals.

A spectrally and spatially dense sensor array consisting of 15 regenerated fiber Bragg gratings (RFBGs) over a length of 30 mm is presented for precise and fast multipoint temperature sensing up to 700 °C. For the first time, it could be shown that with a dense fiber Bragg grating (FBG)-based sensor array the accuracy requirements of Class 1 thermocouples could be achieved and even exceeded. This also represents the highest spatial density of FBG-based high-temperature multipoint sensing reported so far. The mitigation of broadband losses during the regeneration process was studied, revealing the advantages of the low broadband loss characteristics of the RFBGs, especially when larger numbers of measuring points are required. Low measurement uncertainties were achieved by a new, improved calibration methodology and by an analysis of interferences of an FBG with the side lobes of spectrally neighboring FBGs as well as their suppression by a suitable arrangement of the Bragg wavelengths within the array and corresponding data processing. The capabilities of this multipoint sensor technique were demonstrated by resolving the temperature profile within the calibration volume of a calibration furnace and by resolving the temporal and spatial temperature gradients within the flame of a Bunsen burner. The results are of great importance for fiber-optic sensing of high-temperature profiles in real-world applications.

Multimode fiber (MMF) sensors have been extensively developed and utilized in various sensing applications for decades. Traditionally, the performance of MMF sensors was improved by conventional methods that focused on structural design and specialty fibers. However, in recent years, the blossom of machine learning techniques has opened up new avenues for enhancing the performance of MMF sensors. Unlike conventional methods, machine learning techniques do not require complex structures or rare specialty fibers, which reduces fabrication difficulties and lowers costs. In this review, we provide an overview of the latest developments in MMF sensors, ranging from conventional methods to those assisted by machine learning. This article begins by categorizing MMF sensors based on their sensing applications, including temperature and strain sensors, displacement sensors, refractive index sensors, curvature sensors, bio/chemical sensors, and other sensors. Their distinct sensor structures and sensing properties are thoroughly reviewed. Subsequently, the machine learning-assisted MMF sensors that have been recently reported are analyzed and categorized into two groups: learning the specklegrams and learning the spectra. The review provides a comprehensive discussion and outlook on MMF sensors, concluding that they are expected to be utilized in a wide range of applications.

New measurement concepts and their further development lead to a broader range of relevant environmental parameters, such as gas concentrations. This paper presents a photoacoustic gas concentration measurement concept that does not require moving elements due to interferometric optics. The theoretical approach promises compensation for environmental influences such as relative humidity and changes in atmospheric pressure. In addition, practical measurements show a linear behavior concerning the gas concentration. Currently, the measurement system has a resolution for NO2 concentrations in the mid ppb range.