This PDF file contains the front matter associated with SPIE Proceedings Volume 12116, including the Title Page, Copyright information, Table of Contents, and Conference Committee list.

There is an on-going need for sensor technologies capable of providing non-contact chemical detection and identification in the defense community. Here, we present the development of a standoff deep ultraviolet (DUV) Raman sensor for the detection of explosive residues. The sensor is based on a solid-state DUV excitation source coupled with a Spatial Heterodyne Spectrometer (SHS) receiver. The sensor is designed to detect Raman signals from a 4 cm² area surface at a 1 m standoff. Detection and identification is achieved by correlating measured Raman signatures with high fidelity library spectra. The DUV excitation enables operation in a solar blind spectral region, leverages v⁴ cross section scaling and resonance enhancement of Raman signatures, and minimizes the impact of sample fluorescence. The SHS receiver provides a ~100× higher etendue than conventional slit-based spectrometers in a compact and rugged form factor, allowing for high performance field use. This work describes the system design and architecture of the Raman sensor prototype. Developmental standoff Raman measurements with the sensor using bulk liquid and solid samples are presented. Traceability to detection at the µg/cm² scale is demonstrated and future improvements to increase system standoff are discussed.

Drop-on-demand and inkjet printing technology continues to be a promising method of producing chemical test standards with scalability and flexibility to allow for inexpensive, high-throughput production of samples. This enabling technique provides precise, accurate and highly reproducible test coupons that mimic the hazardous chemicals encountered in various theater scenarios; critical in assessing the performance of existing and future sensors detection capabilities. Under the U.S. Army Forensics Advanced Research Program, the Spectroscopy Branch within the Research and Technology Directorate, DEVCOM CBC, along with internal and external collaborative partners are currently utilizing the Direct Color Systems 1800z flat-bed inkjet printer for deposition of various chemicals on relevant surfaces and GeSiM NP2.1 Nanoplotter for more precise and control droplet deposition to support various optical and non-optical detection objectives. The samples produced under this project are used for the evaluation of trace level energetic materials and illicit drugs of abuse within latent fingerprints, deposition of sorbent polymers onto photonic integrated circuits for vapor detection, point sensors, and more recently exploring enhanced training aids for military working dogs. This work will present results from the characterization of utilized chemical deposition techniques as well as recent experimental results from various assessed detection technologies

A proof-of-concept of a scanning stand-off dual-comb spectrometer for explosives detection and identification at 3 m distance is demonstrated. Detection of two types of explosives: RDX and PETN on various surfaces was carried out in reflection-absorption and backscattering modes. A scanning area of 18 cm X 18 cm (400 pixels) was covered in ~2.5 sec. Identification method was based on Pearson’s correlation coefficients between the recorded reflection-absorption (backscatter reflection) spectra and transmission (reflection) FTIR of substances, with baseline subtraction using the asymmetric least square smoothing algorithm. Detection limits of the laboratory system of < 2 ug/cm² were achieved.

Under the U.S. Army Forensics Advanced Research Program, the The Portable Chemical Fingerprint Identification System (P-CFIS) is being developed by U.S. Army Combat Capabilities Development Command –Chemical Biological Center (DEVCOM-CBC) enabling trace level (non-visual) detection of solid particulate contamination on surfaces and residual fingerprints. This operationally flexible non-contact, non-destructive methodology is reducing the need to transport potentially contaminated materials and provide a field forward detection capability yielding greater situational awareness of the threat environment. This uniquely developed prototype based on Raman spectroscopy allow the system to scan a one-inch square area of interest, which may be flat or uneven, target any found particles in the field of view, and automatically analyze and report detection events of threat chemicals contained within the spectral database of the system. This presentation will describe the preliminary results from evaluating the performance of the prototype systems, discussing time of analysis, particle size characteristics, analysis of heterogeneous surfaces, and future development of the systems ultimately leading to the development of the next generation of expeditionary systems for military forensic analysis, checkpoint detection, and/or sensitive site exploitation.

Understanding a system’s performance while operating under different scenarios is difficult because of the vast number of varying parameters that need to be accounted for. To mitigate some of the difficulty a model can be developed that provides some predictability in a system’s performance thereby reducing material usage and laboratory time. It is therefore prudent to understand these parameters and capture that information in order to increase the predictability of a system, especially prior fielding. Through modeling, we connect laboratory scale data with potential scenarios in the field to accomplish this. In this paper, we show that through the modeling of a combination of spectra and instrument operating characteristics we can provide a predictive capability of a system’s performance. Our anomaly detection algorithm can predict a limit of anomaly detection (LOAD) for potential scenarios and then compare them to actual data for validation of our predictive capability. We show similar LOADs in both simulation and actual data collected. We further develop our model to account for realistic field scenarios and evaluate changes in performance.

The optical constants, namely the real (n) and imaginary (k) parts of the complex refractive index, are of interest to generate the infrared (IR) spectra of liquid and solid materials in different morphologies. To obtain n/k, however, most materials are typically not found in the monolithic forms necessary to easily measure n/k, and thus require the use of other methods such as pressing powders to form planar, specularly-reflective pellets. In this work dolomite crystals are measured using fixed-angle IR reflectance spectroscopy in both 1) a monolithic form with the crystal fixed in epoxy and polished, and 2) a pressed-pellet form of the powder of the same mineral. First results for the two methods are compared. It was found that the measured reflectance can vary by as much as a factor of two between the dolomite crystals and the pressed powder forms. For the two-sample preparation approaches the preliminary spectra are compared and the implications and limitations of each method for determining the optical constants of given materials are discussed. Comparison to literature data suggest that polarization effects likely account for the differing amplitude results for reflectance (and hence the k-vectors): Dolomite is biaxial with significantly differing optical constants for the ordinary and extraordinary rays; the pressed pellet method measures an ensemble of microcrystals in randomly oriented positions, whereas the single crystal maintains just one orientation relative to the optical axis.

Variable angle spectroscopic ellipsometry and single-angle infrared reflectance methods have been used to derive the complex optical constants n and k in the mid-infrared spectral region for aspartame. Aspartame exists in four unique forms: three of which incorporate water into the lattice, as well as the anhydrate form. The different forms can induce splitting or slight wavenumber shifts in the spectral features. Pressed pellets of neat powder were prepared and measured using both methods to derive the optical constants. Different n and k values were obtained depending on the percentage of the forms of aspartame, which was determined using powder X-ray diffraction (XRD) analysis.

Spectra of the optical constants (n, k) of a substance are often obtained by comparing spectroscopic measurements of a bulk sample with a simulation model. The reflectance method requires a sample with a perfectly smooth surface to give unbiased values of n and k. The ellipsometric method generates n and k spectra which are accurate in general but which sometimes generate errors over limited spectral ranges when simulating polarized reflectance. We propose a new hybrid method to calculate reliable spectra of n and k. The proposed method uses both ellipsometric and s-polarized reflectance measurements and takes into account the potential roughness of the sample’s surface with the help of a specularity factor. The proposed method provides n and k that better simulate polarized reflectance measurements and applies to isotropic bulk samples with either smooth or rough flat surfaces. We provide demonstrations in the infrared spectral region with a smooth sample and a rough sample.

Analytical technologies including infrared, Raman, and X-ray fluorescence have seeing increasing in-field application over the last 15 years due primarily to advances in miniaturization and on-board embedded analytics focused on giving actionable answers to non-scientist operators. Though a long-time favorite in the analytical laboratory, mass spectrometry has taken longer to attain widespread field adoption largely due to challenges surrounding device portability. In the last few years, mass spectrometry has evolved to a truly handheld state and has continued to broaden its field capabilities and is now seeing significant in-field use as a result. In this presentation we will review state of the art in microscale ion trap systems and present test data from a variety of CBRNE relevant examples, including results generated using a sampling module designed to detect aerosolized threats

Robust sample collection and identification of trace amounts of compounds outside of a laboratory environment is a challenge facing military users, first responders, and law enforcement An increasing number of portable instruments are being developed to focus on improving on-site sampling and analysis. A weighty obstacle for these field-focused systems is the ability to detect trace amounts of analytes from complex matrices. Previous work has shown the benefits of utilizing pressure-sensitive adhesive (PSA) coated paper for collection combined with paper spray ionization mass spectrometry (PS-MS) for the identification of trace amounts of small molecule. In this work, analysis of explosives captured on a PSA substrate via a portable instrument was examined. Positive identification was achieved for TNT, HMX, and RDX when sampling from surfaces containing less than 1 mg of each explosive. It was also determined that mixtures of the explosives and illicit drugs could be identified even with the presence of interferents. Additional experiments were performed to extract and reanalyze substrates, as well as longevity stability studies. The results demonstrate the potential value that PSA substrates combined with PS-MS can provide in a field forward or first responder setting.

While infrared spectroscopy is often used for chemical vapor identification, it has two major disadvantages: relatively low sensitivity and the inability to reliably identify components in a complex mixture. This paper presents a method that overcomes the low sensitivity challenge. The low-sensitivity is overcome by using a thin germanium wafer in an ATR configuration, providing 10 to 20 times more bounces per centimeter than commercially available multi-bounce ATR crystals. By using a 25 mm long crystal, we can detect nanogram amounts of analyte which is adequate for possible applications in detecting CWA threats, environmental monitoring and medicine (e.g. breath analysis). While not specifically discussed in this paper, the mixture challenge can be overcome by using a “cocktail problem” algorithm which requires multiple spectra with differing concentration ratios of analytes. This is provided by collecting the sampled vapor on either a single spin-coated sorbent film layer or several parallel sorbent strips with differing chemistries to separate the mixture into classes. In both cases, temperature ramping of the sorbent-coated crystal provides additional unique spectral data through boiling point separation. Full gas chromatography column separation is also possible and this is the main topic of a companion paper. In the current experimental setup, we use a tunable quantum cascade laser as the light source and a TE cooled MCT detector.

Infrared technology can provide a wealth of information related to biological and chemical hazards in the environment. However, this technology mostly exists in the form of bulky instrumentation on optical benches in academic laboratories. We discuss the transition of IR sensing to various points-of-need applications, including food and water safety, bioreactor process control and chemical analysis of drinking water. In particular, in remote locations the access to clean drinking water is critical to soldiers’ health. Mid-infrared spectroscopy is a powerful tool for identification and quantification of a wide range of common organic and inorganic compounds. In this contribution we present data demonstrating proof-of-concept of a quantum cascade laser (QCL)-based infrared sensor for evaluation of toxic industrial chemicals (TICs) and toxic industrial materials (TIMs) and discuss the path for development of miniaturized, point-of-need IR photonic integrated circuits (IR-PIC).

Polymer sorbents able to selectively capture specific classes of analytes have attracted significant interest in the context of photonic sensors (particularly waveguide-enhanced Raman spectroscopy, or WERS) for chemical warfare agents and industrial gases. We have developed a method for using aminopropyl methylsiloxane-dimethylsiloxane copolymers as inexpensive starting materials for the synthesis of new sorbents for chemical and biochemical sensing. Conversion of the starting polymer to a product sorbent can be accomplished via a simple, single-step reductive amination reaction with an aldehyde. Preliminary tests of a sorbent in the context of refractive index-based sensing of energetic compounds (explosives) using silicon nitride micoring resonators is also discussed.

The concentration of small molecule biomarkers in human serum and saliva has been shown to be characteristic of viral disease and correlated with disease severity. Inexpensive point-of-care diagnostic methods to quantify and track these analytes would provide additional information beyond viral or antibody detection assays to guide diagnosis and therapy. Waveguide-enhanced Raman spectroscopy (WERS) enables the detection and identification of trace concentrations of dissolved analytes using a chip-scale photonic circuit based on long evanescent waveguides. Here, we describe WERS measurements of two biomarkers: glucose and urea. This proof-of-concept work will provide the basis for the development of handheld bio-marker detection systems based on packaged photonic circuits integrated with a laser source and detector.

Chemical Biological Radiological Nuclear and Explosive (CBRNE) sensing systems in the field provide alarms in the form of simple graphical representations, lights, vibrations, and alarm sounds to maximize the reaction time of the user in the event of a hazardous situation. Artificial Intelligence (AI) can be used to reduce the false alarms of chemical detectors, allowing users to react with confidence when an alarm does occurs. However, the Department of Defense’s AI ethics standards states that technologies incorporating AI systems be traceable, reliable, and governable. Given the complex nature of AI and the difficulties of interpreting results, testing and evaluating AI systems poses a challenge for CBRNE sensing systems. To properly interpret and evaluate AI systems it is imperative graphical user interfaces (GUI) are designed to be simple interfaces that provide easy to interpret results. Presented here is an interpretable alarm GUI for orthogonal networked sensors (IAGOnet). IAGOnet provides real-time status of connected sensors utilizing a familiar replication of their onboard results, along with simple to understand graphical representations of confidence metrics from machine learning (ML) predictions. IAGOnet allows a user to compare the detector’s original alarm state to current and previous predictions of classification algorithms, thereby reducing the false alarms. Our work demonstrates the practical nature of IAGOnet by utilizing data from an ion mobility spectrometry (IMS) based detector and a multi-gas detector.

The accidental release of industrial and agricultural chemicals can pose a serious threat to life and the environment. Therefore, researchers have been exploring detection methods of commonly transported chemicals in order to minimize potential harm or destruction in response to an accidental release. One method is to use a network of commercial sensors to track a chemical spill but with each sensor costing upwards of $600, this type of network can become prohibitively expensive and may not be practical for real world use. Specifically, we aimed to develop a network of custom electrical conductivity sensors with each sensor made from an inexpensive Arduino board showing comparable detection results while costing an order of magnitude less. In our experiments, the network of sensors covered 83 in2 in a container filled with different types of water (e.g. deionized, melted snow, sea, river, and tap). The network of custom sensors showed high ammonia concentrations near the release point of an initial laboratory scale ammonia release with low ammonia concentrations away from the release point. As equilibrium was reached, the sensors showed the same ammonia reading. Additionally, a 2-D map was made to track the simulated ammonia spill overtime. Overall, this works shows that this network of custom Arduino sensors can be used to map the detection of accidental ammonia release as an inexpensive replacement for the commercial sensors, which will promote accessibility of future testing for the broader community.

We describe here a device for the autonomous use of Draeger colorimetric sensor tubes (CSTs) at large stand-off distances (< 1 km). CSTs are an attractive option for the detection of hazardous chemicals thanks to their high sensitivity and simplicity but present several challenges when considered for automation. Here, these challenges are discussed, and a proof-of-concept device that demonstrates viable solutions is detailed. Our realized prototype is capable of processing thioether CSTs without human interaction and demonstrates comparable sensor fidelity to manually processed tubes.

Artificial illumination is required for a line scanning passive hyperspectral spectrometer when operating a system of this type in non-daylight conditions. While in general more photons will yield a larger reflectance signal return to the sensor, a source that outputs a large number of photons is unlikely to be compatible with a compact hyperspectral spectrometer on a small aircraft or using in a handheld manner. Therefore, in this paper we investigate a small tungsten halogen source coupled with off-the-shelf optics to create a compact artificial illumination source to provide photons for the spectrometer. After characterizing the compact halogen source and comparing its output characteristics to larger sources currently in use, several optical trains were designed to focus the sources output to the instruments’ field-of-view. The results detailed herein show that a compact source can allow for a hyperspectral spectrometer to operate with a compact artificial illumination source with a minimal decrease in performance.

The present work details how convolutional and recurrent deep learning networks can be used to classify infrasonic and seismic events of regional-field rocket demolitions, rocket motor burns, quarry blasts, and earthquakes. To accelerate machine learning adoption within the geophysical sciences, we illustrate the full machine learning pipeline: data acquisition and cleaning, preprocessing, model construction and training, and model understanding using feature-space analysis. Multiple deep learning architectures are evaluated to provide practical lessons learned and insights. The LSTM-RNN suffers from degraded learning on long geophysical signals, while a CNN is more robust to variance in data quality and length. Geophysical time-series signals should be learned with the instrument response deconvolved to avoid gross resampling or decimation of the signal. Frequency-domain feature inputs like spectrograms exhibit improved classification performance, and mapping event-based attributes to the learned feature space of deep networks can provide explainable physical context. All network configurations are validated on geophysical data collected in the Utah region, with experimental results including ablation studies examining different input types, preprocessing strategies, and hyperparameter settings.

The MAESTRO instrument is a recent development that enables sensitive, spatially located, microbial detection using deep UV spectroscopy at a standoff distance up to 5m. This capability stems from prior deep UV fluorescence/Raman standoff instruments for point chemical, biological, and explosives analysis, as well as from planetary science in the form of the SHERLOC instrument on the Mars 2020/Perseverance Rover, a deep UV fluorescence/Raman mapping robotic arm-mounted instrument looking for signs of life on Mars. The MAESTRO instrument leverages these detection capabilities to enable microbial detection on environmental/natural surfaces with high-speed mapping rates with map areas <1m2 . This talk will discuss the fundamentals of the methodology for detection, the achieved sensitivity, and the analytical approach used to detect and differentiate microbial hazards from the background.

There is a need for time efficient evaluation methods to discriminate between viable and dead bacterial spores. In this work, the potential to use the autofluorescence from spore suspensions for evaluation of spore deactivation processes is investigated. Bacillus thuringiensis and Bacillus anthracis ATCC 4229 spores were exposed to UV-radiation for deactivation and the fluorescence response was monitored at different radiation doses and the deactivation was evaluated via traditional bacterial incubation on agar culture plates. For excitation wavelengths of, e.g., 280 m and 330 nm, differences in the fluorescence response could be observed for different live:dead ratios.

Here we describe the performance of a recently developed infrared-based chemical sensor for detecting liquid chemical hazards on surfaces at stand-off distances. This sensor uses the signal acquired through three broadband infrared optical filters in the mid and long wave infrared (MWIR LWIR), as opposed to requiring a spectrally tuneable source (such as Quantum Cascade Laser systems) or an FTIR spectrometer, to discriminate between target chemicals and background interferents. Although this technique doesn’t interrogate the full MWIR/LWIR spectrum, it’s ideal for screening a pre determined list of hazards as well as lowering system cost and complexity. The technique is the IR analogue of how the human eye discriminates between different colors, and utilizes modified International Commission on Illumination (CIE) chromaticity charts to illustrate chemical detection performance. We demonstrate that the sensor can correctly discriminate between the chemical warfare agents VX and T-mustard, as well as DEET (N, N-Diethyl-meta-toluamide) insect repellent deposited upon surfaces at a distance of 1.9 meters.

This paper describes the development status of a standoff handheld sensor for real-time assessment of trace chemical, biological, and explosives materials on surfaces, which is an important capability needed by warfighters/first responders for situational awareness of hazards in their proximity. It is important to perform these assessments without contact or spreading of substances or use of reagents. This work was conducted under Army, DTRA, and DHS funding to develop a Standoff Hand-held CBE (SHCBE) optical sensor which detects and classifies trace and bulk concentrations of a wide range of CBE materials on surfaces at distances of 1 to 5 m in real-time and full daylight with a fully integrated analyzer. The sensor method combines independent but complementary chemical information of molecular bonds within a targeted material using Raman and electronic configuration information of the material using fluorescence, both with excitation below 250 nm. There are seven primary advantages of the SHCBE detection method compared to near-UV, visible or near-IR counterparts: 1) Solar blind detection enabling standoff operation in full daylight; 2) Fluorescence-free Raman and Raman-free fluorescence enabling enhanced detection and identification of target materials without mutual interference; 3) Resonance Raman signal enhancement for improved Raman sensitivity; 4) Simplification of Raman spectra due to resonance enhancement, 5) Short penetration depth, providing physical separation of surface contaminant materials from substrate; 6) no damage to sensitive organic and biological materials, due to laser pulse width related sample heating, and 7) eye retina safe. These capabilities are not possible with near UV, visible, or near IR sensors. A special feature of our sensor is the ability to detect trace biological materials at standoff distances in real time with a handheld device. Photon Systems has developed these methods over many years, enabling instruments deployed to extreme environments on Earth and the SHERLOC instrument which has been successfully operating on Mars on the Perseverance Rover since it landed on Feb. 18, 2021

Alakai Defense Systems has recently developed what we believe is the first one-handed UV Raman sensor for standoff detection of chemicals, which we refer to as the Situational Awareness for First Responders (SAFR-1). Alakai is now expanding that product line to offer a higher performance version called SAFR-2 for longer range and trace detection. SAFR-1 detects bulk and residue quantities of material up to a range of 5 m. Since it is lightweight, SAFR can also be deployed on Unmanned Ground Vehicles (UGV’s) or Unmanned Aerial Vehicles (UAVs). SAFR-2 incorporates an Intensified CCD (ICCD) array for more sensitivity and trace detection capability.

We present the further development of a cart-based system for infrared backscatter imaging spectroscopy (IBIS) designed to detect and analyze trace amounts of hazardous materials at proximal stand-off distances. A four-chip quantum cascade laser system quickly scans through the mid- to long-wave infrared (6 µm – 11 µm) wavelength range to illuminate samples contaminated with analyte. The backscattered light from the targets is collected with a liquid nitrogen cooled MCT focal plane array. Wavelengths are assigned to each frame collected with the MCT camera corresponding to the emission of the laser at the time of acquisition. This process builds a hyperspectral image cube containing spectral reflectance data for every pixel in the image. The experimental results of this cart-based infrared illumination and backscatter detection are presented. A single detection event can be completed in less than 1 second, and every pixel of the 128x128 camera array produces an individual spectrum. Advancements in this setup include mitigation of QCL beam wander and differentiating between nine analytes all present within the same one square inch target. Reference spectra of the target analytes are measured using a high resolution FTIR to validate the highly sensitive and chemically specific nature of the IBIS cart-based measurement. The sample was prepared to mimic real-world threats such as explosives and illicit drugs in trace amounts on relevant substrates.

A novel multi-path extinction detector (M-PED) is being developed for point detection, identification and quantification of vapor phase chemicals. M-PED functions by pairing a broadband long-wave infrared (LWIR) quantum cascade laser with a novel sample cell, designed to simultaneously measure chemical absorption at multiple pathlengths and wavelengths. The pathlength samples are angularly separated in one dimension, such that a diffraction grating can be used to measure wavelength data in the orthogonal dimension using a compact, low-cost microbolometer array. The resulting data matrix is fit to Beer’s Law in two dimensions to accurately quantify chemical concentration while rejecting common mode noise (e.g. laser amplitude noise). The design, characterization and a capability demonstration of the advanced prototype sensor are presented.

Standoff detection of chemicals remains a crucial need for a variety of applications of importance for defense, homeland security, environmental, and industrial applications. The goal of standoff chemical sensing is to enable the identification and classification of an unknown hazardous or toxic chemical, without any operator or instrument having to come in direct contact with the chemical itself. Currently, standoff detection of chemical vapors is carried out using optical sensing techniques. Passive infrared (IR) sensors have identified chemical vapor clouds at ranges exceeding one kilometer by detecting, spectrally resolving, and analyzing scene radiance. Currently available passive IR sensors have substantial size, weight, power, and cost (SWaP-C) limitations, which reduce the number of sensors capable of being deployed in a given area, or precludes their use altogether in certain circumstances. To address these limitations, we are developing a unique passive low SWaP-C IR sensor capable of detecting chemical vapors when viewed against a cold-sky or terrestrial background. This sensor, inspired by human color vision, will use only the response through three broadband infrared optical filters to discriminate between target chemicals and background interferents. The key technology of the PBS is a commercially available pyroelectric quadrant chip sensor which contains four channels with unique bandpass IR filters installed. We demonstrate results collected using a variable temperature blackbody in the laboratory, which represents passive IR sensing against various background conditions. These results demonstrate the first step in the development of a passive bioinspired IR sensor which will use only low-cost commercially available components, and be capable of rapidly providing actionable detection of chemical vapor clouds

We present further development of an eye-safe, invisible, stand-off technique designed for the detection of target chemicals (such as explosives) in a single “snapshot” frame. Broadband Fabry-Perot quantum cascade lasers (FP-QCLs) are employed as active illumination sources, in the Mid-LWIR (long-wave infrared) in the range of 7 to 12 µm, to interrogate the spectral features from analytes of interest. We have developed a custom-built broadband laser source utilizing an OEM FP-QCL. This “white” broadband laser source enables stand-off detection in a single snapshot frame. Light from this source was collimated and aligned toward the target several meters away. The “backscatter” and absorption signals from target chemicals are spectrally extracted by an LWIR spectrometer based on the spatial heterodyne spectroscopy (SHS) technique. The SHS offers high throughput and full spectral coverage in each single frame from an IR imaging array. This manuscript will cover the implementation and optimization of FP-QCLs for this broadband spectroscopic application. We will also discuss the operation and processing of SHS images to extract spectral information. Finally, we will present results of measurements using specific analytes to demonstrate the application of the method to stand-off detection of targets such as explosives and other chemical threats.

Carbon dioxide (CO₂) is an atmospheric trace gas, and its accurate sensing is therefore of great interest. Optical sensors exploiting the mid-infrared (mid-IR) light absorption of CO₂ provide high sensitivity and are widely used in medical diagnostics, atmospheric monitoring, remote sensing, and industrial applications. In this work, we demonstrated an accurate CO₂ gas sensing at 4.2 µm wavelength. In addition, detecting the weak mid-IR molecular absorption bands of gases at low concentrations requires increasing optical path lengths to be used. The most obvious method that can expand the potential beam path in a spectroscopic system is to use a longer linear gas cell, which in some situations may be adequate; however, space and volume requirements need to be considered. In this work, a circular multi reflection (CMR) cell was used to reflect the radiation back and forth through the sample medium multiple times, thus greatly reducing the size footprint compared to a linear cell of equivalent optical path length. A CMR cell was designed and constructed to allow multi-reflections within the cell. The optical alignment of the cell and the convenience of changing the optical path length by adjusting its position with respect to the incident light beam were also used to maximize the advantages of the device. This work will be used as the groundwork for designing an instrument for the high-resolution measurement of CO₂ gas in planetary atmospheres.

In this paper, we discuss the characterization of several quantum cascade lasers (QCL) as candidates for a broadband infrared illumination source for use in single “snapshot” detection of hazardous materials. Each of the lasers discussed is a Fabry-Perot quantum cascade laser (FP-QCL) chosen for its peak emission within the mid- to long-wave infrared region of 7 µm to 12 µm. These lasers are commercially available from several vendors. The output of each laser was characterized using a high resolution FTIR spectrometer to record each laser’s emission spectrum under varying operating conditions such as driving current, QCL temperature, and operating modes (continuous wave or pulsed). Time-Resolved Spectroscopy (TRS) was performed on each laser’s pulsed driven output to provide further details on how the emission of each laser evolves on the nanosecond time scale. We specifically investigate and present spectra of FP-QCL packaged in sealed OEM configurations. These devices offer center wavelengths ranging from 8.9 µm to 10.5 µm. We present the results of changing operating conditions to optimize the QCL emission to provide high-power and broad spectral coverage. By combining two or more FP-QCL, we obtain spectral coverage of approximately 3 µm. The purpose of this study is to develop a high-power, broadband, “white light” illumination source to provide wide spectral coverage over the region of interest for standoff detection and analysis of potentially hazardous materials.

The ability to rapidly detect hazardous airborne chemicals in a complex chemical background with high fidelity remains a significant challenge. Separation through traditional Gas chromatography (GC) can significantly augment most detection technologies for high fidelity detection, but with the disadvantage of requiring the chemicals to elute off the column before detection can occur. This translates to added time for any decision-making process. Microfabrication of GC systems has reduced footprint and power consumption, but the end-of-column detection paradigm has remained. We present the first in-column detection system which probes the GC stationary phase, coated on an IR transparent column substrate, with an active infrared source. The optical evanescent field interacting with the stationary phase (US. Patent# 9,599,567, Navy Case number 211024-US1) allows for detection along the column without having to wait for complete elution. These spectral signatures, collected at different regions along the column, are analyzed by an algorithm to identify components in a complex mixture. We present results with an ATR-based system with a molded micro-GC column whose base comprises an optically transparent material coated with the stationary phase on proof of concept mixtures.

Datasets with accurate ground truth from unmanned aerial vehicles (UAV) are cost and time prohibitive. This is a problem as most modern machine learning (ML) algorithms are based on supervised learning and require large and diverse well-annotated datasets. As a result, new creative ideas are needed to drive innovation in robust and trustworthy artificial intelligence (AI) / ML. Herein, we use the Unreal Engine (UE) to generate simulated visual spectrum imagery for explosive hazard detection (EHD) with corresponding pixel-level labels, UAV metadata, and environment metadata. We also have access to a relatively small set of real world EH data with less precise ground truth – axis aligned bounding box labels – and sparse metadata. In this article, we train a lightweight, real-time, pixel-level EHD pre-screener for a low-altitude UAV. Specifically, we focus on training with respect to different combinations of simulated and real data. Encouraging preliminary results are provided relative to real world EH data. Our findings suggest that while simulated data can be used to augment limited volume and variety real world data, it could perhaps be sufficient by itself to train an EHD pre-screener.

The objective of this paper is to present an efficient parallel implementation of the iterative compact high-order approximation numerical solver for 3D Helmholtz equation on multicore computers. The high-order parallel iterative algorithm is built upon a combination of a Krylov subspace-type method with a direct parallel Fast Fourier transform (FFT) type preconditioner from the authors’ previous work, as shown in Ref. 7. In this paper, we will be presenting the result of our algorithm by computationally simulating data with realistic ranges of parameters in soil and mine-like targets. Our algorithm will also be incorporating second, fourth, and sixth-order compact finite difference schemes. The accuracy and result of the fourth and sixth-order compact approximation will be shown alongside the scalability of our implementation in the parallel programming environment.

We investigated the utility of target visibility and dissimilarity metrics proposed by Pezzaniti et al.? Metric utility was determined by looking for correlation between the metrics and ML object detection performance. To do this, thousands of synthetic image sets were generated with varied seed parameters, such as target mean, background mean, target contrast and background contrast. The visibility and dissimilarity metrics for the synthetic images were calculated. YOLOv5? was used to detect simulated military threats in the synthetic data. We will present correlations to determine the significance of the visibility and dissimilarity metrics as it relates to detection performance.

The ability to detect and classify buried objects using thermal infrared imaging is affected by the environmental conditions at the time of imaging, which leads to an inconsistent probability of detection. For example, periods of dense overcast or recent precipitation events result in the suppression of the soil temperature difference between the buried object and soil, thus preventing detection. This work introduces an environmentally informed framework to reduce the false alarm rate in the classification of regions of interest (ROIs) in thermal IR images containing buried objects. Using a dataset that consists of thermal images containing buried objects paired with the corresponding environmental and meteorological conditions, we employ a machine learning approach to determine which environmental conditions are the most impactful on the visibility of the buried objects. We find the key environmental conditions include incoming short-wave solar radiation, soil volumetric water content, and average air temperature. For each image, ROIs are computed using a computer vision approach and these ROIs are coupled with the most important environmental conditions to form the input for the classification algorithm. The environmentally informed classification algorithm produces a decision on whether the ROI contains a buried object by simultaneously learning on the ROIs with a classification neural network and on the environmental data using a tabular neural network. On a given set of ROIs, we have shown that the environmentally informed classification approach improves the detection of buried objects within the ROIs.

Recent advancements in signal processing and computer vision are largely due to machine learning (ML). While exciting, the reality is that most modern ML approaches are based on supervised learning and require large and diverse collections of well annotated data. Furthermore, top performing ML models are black (opaque) versus glass (transparent) boxes. It is not clear what they are doing and when/where they work. Herein, we use modern video game engine technology to better understand and help create improved ML solutions by confronting the real world annotated data bottleneck problem. Specifically, we discuss a procedural environment and dataset collection process in the Unreal Engine (UE) for explosive hazard detection (EHD). This process is driven by the underlying variables impacting EHD: object, environment, and platform/sensor (low altitude drone herein). Furthermore, we outline a process for generating data at different levels of visual abstraction to train ML algorithms, encourage improved features, and evaluate ML model generalizability. Encouraging preliminary results and insights are provided relative to simulated aerial EHD experiments.

This paper introduces a direct parallel partial FFT-type algorithm for the numerical solutions of the two- and three-dimensional Helmholtz equations. The governing equations are discretized by high-order compact finite difference methods. The resulting discretized system is indefinite, making the convergence of most iterative methods deteriorate as frequency increases. In this situation, the parallel direct approaches are a better alternative, especially for the systems with discontinuous and singular right-hand sides. The research focuses on the efficient parallel implementation of the proposed algorithm in shared memory environments (OpenMP). The complexity and scalability of the direct parallel method are investigated on scattering problems with realistic ranges of parameters in soil and mine-like targets.

In this contribution, the performance of a Ground Penetrating Radar (GPR) system mounted on board an Unmanned Aerial Vehicle (UAV) to detect buried landmines and Improvised Explosive Devices (IEDs) is analyzed. Radar measurements are coherently combined using a Synthetic Aperture Radar (SAR) algorithm complemented with several clutter mitigation techniques. As a result, 3D high-resolution radar images of the subsurface with enhanced signal-to-clutter-ratio are retrieved. Several prototypes and different GPR architectures have been extensively tested in realistic scenarios, where numerous metallic and non-metallic targets (landmines and IEDs) have been buried under different conditions (dry and wet fields, dirt roads, sloped terrains).

With the increasing use of RF jammers, detection of buried command wires has gained significant importance in detecting metallic and non-metallic threats which may be buried underground. The concept of using a multi-polarimetric Ground Penetrating Radar (GPR) is demonstrated for the purposes of detecting command wires. Considering the wire diameter, orientation and the buried depth the need for high resolution for wire detection increases even more. For this reason, L – C band Ultra-Wide Band (UWB) multi-polarization antenna GPR antenna array is studied. Using a full-wave electromagnetic simulation environment, which includes the models of Quad Ridge Horn (QRH) antennas, clayey soil and the target wire, the problem is modelled and the obtained performance is demonstrated. Furthermore, measurements were taken with an experimental GPR data collection setup in a similar fashion to the full-wave electromagnetic simulation environment.

Wire detection is often based on line structure. This research investigates the use of ellipse feature to detect surface laid and shallowly buried wires. A wire has non-negligible diameter and elliptical shape appears in its cross-section image. After filtering and edge detection in the cross-section image, ellipse fitting is applied to obtain an ellipse feature for indicating how well the shape fits to an ellipse. Experimental results of the ellipse fitting technique are presented for the detection of wires and their discrimination with clutter objects.

Detecting and mapping underground utilities remains a problem in the United States. Poor and outdated utility maps can lead to difficulty locating buried infrastructure. Aging and poorly maintained utility infrastructure can leak or fail, endangering public safety and the environment. Many geophysical sensing techniques have been used to locate and map buried infrastructure, including: acoustic methods, ground penetrating radar, passive magnetic fields, and low frequency electromagnetic fields – each with its own advantages and pitfalls. Our previous work has focused on exciting linear currents on PEC thin wires using 100-1000 kHz electric fields and observing the secondary magnetic field response of these wires. This paper extends that work and investigates the wideband electromagnetic signature of pipes in the ground. The effects of pipe material and inhomogeneity are examined by using the Method of Moments (MoM) and the Method of Auxiliary Sources (MAS). The surface-impedance boundary condition (SIBC) is used to account for the finite conductivity of the pipe.

This paper describes an active magnetic sensing system that locates and distinguishes buried ferromagnetic and non ferromagnetic objects. The operating principle is that primary active magnetic fields interact with ferromagnetic and non ferromagnetic conductive materials to produce secondary fields that can be measured and interpreted. Ferromagnetic materials produce a secondary magnetic field by induced magnetization. Conductive materials produce secondary magnetic fields with eddy currents that counter the primary field. The system uses a compact array of four electronically geared, rotating multipole neodymium magnets to project the primary shape-controlled oscillating magnetic fields. Magnetometers measure the combined primary and secondary fields at a rate of 490Hz. The data are then read into a mini PC to characterize in near real-time the composition of subterranean objects.

A rapid, portable, and cost-effective method to detect the infection of SARS-CoV-2 is fundamental toward mitigating the current COVID-19 pandemic. A localized surface plasmon resonance (LSPR) sensor based on human angiotensin converting enzyme 2 protein (ACE2) functionalized silver nanotriangle array is developed for rapid coronavirus detection. The sensor is validated by SARS-CoV-2 spike RBD protein and CoV NL63 virus with high sensitivity and specificity. A linear shift of the LSPR wavelength and transmission intensity at a fixed wavelength (750 nm) versus the logarithm of the concentration of the spike RBD protein and CoV NL63 is observed. The limits of detection for the spike RBD protein, CoV NL63 in untreated saliva are determined to be 0.38 pM, and 625 PFU/mL, respectively, while the detection time is found to be less than 20 min. Such a LSPR sensor could serve as a potential rapid point-of-care diagnostic platform for COVID-19.

Mid-infrared laser-based sensors are commonly used to detect and quantify many chemical species for environmental, industrial, defense, and security applications. Data-driven approaches, including machine learning and information theory, can be applied to photonics-based sensors to quantify drifts and improve precision. These methods are used to classify signals from rotational-vibrational absorption spectra of Nitrous oxide (N2O) in the 4.3 m region of the spectrum. The detection method utilizes the structural complexity of wavelength modulation spectroscopy signals and information encoded in the spectra. We create our basic training models by simulating temperature, pressure, density fluctuations effects, and molecular transition line broadening of a Voigt lineshape profile. Instrument (laser and detector) noise optical fringing effects can be incorporated in the models. The paper shows that signal variations due to Trace gas density fluctuations and molecular collision dynamics can be discriminated from instrument drifts. The proposed methodology can be used to accurately predict, detect, and evaluate short-term and long-term drifts in sensing systems which can be integrated with the conventional Allan variance methods. We demonstrate this methodology by high-precision sensing of rotational-vibrational transitions of Nitrous oxide and carbon monoxide using an interband cascade laser operating at a 4.3 m spectral region.

Several chemical warfare chemicals have fingerprint spectral signatures in the mid-infrared region of the spectrum. For instance, Sarin is one of the most lethal warfare agents that is a highly toxic synthetic chemical organophos phorus compound. Due to complex chemical structure and large absorption and collision cross-section, the molecular linewidths of such chemicals can cover a broad range of spectral width. Detection of such molecules in the mid-infrared region is sensitive which requires broadly tunable sources and detection methods. We show a rapid detection methodology of such chemicals using proxy methane and nitrous oxide atmospheric bands in the 7 µm to 8 µm region which also have fingerprints region of several hazardous chemicals. Methane absorbs strongly in the wavelength range of 3 µm to 8 µm, nitrous oxide has absorption from 5 µm to 8 µm. As the large wavelength range that they have covered, we use molecular rotational-vibrational transitions of CH4

Inverse spectral analysis of diffuse reflectance for surface-distributed material particles on substrates is described. Inverse spectral analysisis applied using a methodology for extraction of target spectral features, which is based on diffuse reflectance theory and phenomenological multiplicative-factor decomposition of reflectance functions. Specifically, this methodology entails feature-extraction using reflectance-spectrum normalization with respect to phenomenological backgrounds. Case-study inverse spectral analyses of diffuse reflectance for surface-distributed caffeine particles on substrates demonstrate the methodology.