Acoustic Sensor Research Articles

ATD-300 an AI empowered miniature acoustic device for automatic gunshot attacks detection and image capture

Built on over 30 years of acoustic source recognition expertise in transient impulsives signature, recently highly empowered by neural networks, and wave propagation model driven algorithms, ACOEM has developed the ATD-300, an acoustic sensor for automatic gunshot detection, identification and localization. What sets this sensor apart from the competition is its reliability in terms of detection rate over false alarm and robust high resolution shooter pointing capability. Therefore it can be used directly by police departments without third-party human validation. Resulting from the design effort to distribute intelligence among a collaborative sensor network, processing and artificial intelligence algorithms are fully embedded locally, enabling an alert to be raised in less than three seconds and a camera to be pointed at the shooter with only one sensor. In addition, as the system is already installed in several American cities, we will show how a device-to-server distributed artificial intelligence approach continuously improves the performance of the solution, fed by real field data.

Neural-network-based sound source localization from unmanned aerial vehicles

As a mobile platform, unmanned aerial vehicles (UAV) can carry optics and acoustics sensors recording visual and audio information from ground. Over the past decade, UAVs are widely used in different applications, e.g., search and rescue, pipe leakage detection, etc. Compared with visual information, audio information is challenging to capture from UAVs due to the strong ego-noise generated by the rotating motors, propellers on the UAV as well as the motion of the UAV, which result in a low signal to noise ratio (SNR) of the received acoustics signals. With the contaminated signals, the performance of sound source localization algorithms, e.g., conventional beamforming, is severely degraded. To improve the quality of sound source localization result, different sound source enhancement algorithms have been proposed by researchers. In this paper, simulation experiments were conducted to study the potential of apply convolutional neural networks (CNN) to improve the beamforming result. It was found that with CNN, the estimation of a single point sound source location and source strength can be improved. The model's localization accuracy and source strength estimation accuracy were evaluated respectively.

Consideration on the efficient calibration method for low-cost microphones

Current microphone calibration methods in the international standards are basically targeting laboratory and working standard microphones specified in the IEC 61094-2 and IEC 61094-4, respectively. These methods are promising to ensure the reliability however there are many difficulties in applying it to low-cost acoustic sensors such as MEMS microphones, which are rapidly becoming widespread, with aspects of time and cost. Moreover, since the installation conditions for these low-cost microphones are different from those of standardized measurement microphones, calibration methods that take these usage conditions into consideration are required. As potential methods to overcome the problems, the feasibility of the active coupler method applicable to surface mounted condition and the bulk calibration method in the reverberation room are investigated. Both proposed methods can be applied under actual installation conditions and enable rapid and cost-efficient calibration. In this study, methods and procedures for securing traceability are proposed, and uncertainty evaluation is conducted.

Construction site environmental pollution management - initiatives, innovation and research

The management of pollution, and in particular of the impact of noise and vibrations, is very important for construction activities, even more so when it comes to urban construction sites. Indeed, some operations have very significant emissive potential and it is appropriate, for the proper conduct of the activities, to adopt a certain strategy in order to control the impacts and thus ensure good acceptance of the construction by the neighborhood. Environmental monitoring is the key to achieving this dual objective: prevent pollution and communicate. This paper presents our innovative approach to localize, identify and report noise pollution on construction sites for efficient noise mitigation. More precisely, usage of acoustic high resolution beamforming is discussed to both identify predominant sources and estimate reliably their contribution to overall noise pollution captured at a given reference point by à Class I metrological device. By combining the easy-to-deploy ATD directional acoustic sensor with its 10cm aperture microphone array, with video recording for event confirmation and labelling on the highly responsive Cadence interface, we show how we can solve complex use cases for a wide range of users, from site managers to acoustic experts.

One-third octave spectrum calculation to meet the attenuation limits of a Class 1 analyzer (IEC 61260-1:2014) for low-power noise sensors

Thanks to the advances in technology and the cost reductions of electronic components, nowadays it is possible to produce low-cost and low-power acoustic sensors with extended capabilities such as the one-third octave band spectrum calculation. However, there are some doubts about the accuracy of these devices. Thus, to meet the attenuation limits of a Class 1 analyzer for low frequencies as stated in IEC 61260-1:2014 normative, this research proposes two algorithms to calculate the one-third octave band spectrum, both algorithms are based on the Power Spectral Density (PSD). The first one consists of a subdivision of frequency bins, where a PSD bin is divided into smaller parts, gain adjusted, and assigned to its corresponding one third-octave band. The other one is based on signal decimation-accumulation, where samples, that are processed every 125 ms due to the sensor design, are first decimated and then accumulated to increase the frequency bin resolution. Both algorithms are programmed into the sensor and the power consumption and acoustical performance are evaluated and compared to a typical multi-rate filter bank. It is observed that both proposed algorithms improve the performance of the low-cost sensor in terms of energy consumption and accuracy.

Pertinence of finite element method for impulse localization in urban environments

Although urban combat is nothing new, recent conflicts have shed light on Western armies' difficulties when operating in such environments. Thus, new surveillance and hazard detection challenges in complex environments are addressed. A new decision aid tool for acoustic sensor positioning has been developed to improve reconnaissance performance. The software's sensor location optimization kernel is based on a genetic algorithm developed by our partners from FKIE. It resolves successive localization problems using pre-computed Times Of Arrival (TOA) between an impulsive acoustic source and various surveillance microphones. The present study discusses the pertinence of the Time-Domain Finite-Element Method (TD-FEM) to simulate wave propagation within an urban area to obtain TOA and resolve localization problems. The approach is compared to experimental measurements in an urban area, and the pros and cons of the TD-FEM for this configuration are discussed.

Development of low-cost MEMS microphone array for sound localization in urban environments

Understanding noise pollution and mitigating its impact on the quality of life in urban environment is a major challenge from a societal and health point of view. In order to monitor and ultimately understand these noisy environments a process of long term acoustic measurements and analysis is required. These urban environments are noisy and obtaining a clear recording of a desired acoustic signal is difficult hence an effective approach is the use of beamforming theory coupled with a microphone array. In this paper a novel approach utilizing compact low-cost microphone arrays equipped with MEMS sensors for continuous sound localization in urban areas is proposed. This approach focuses on leveraging the affordability and scalability of MEMS sensors to deploy distributed acoustic sensor networks across diverse urban locations. The aim is to capture real-time acoustic data, enabling comprehensive monitoring of sound sources and their spatial distribution within the urban environment. This sensor unit can be upgraded with a camera, i.e. acoustic camera, which will provide information on the conditions of the environment where the measurements are performed. This low-cost solution facilitates long-term monitoring, providing valuable insights into the dynamic nature of urban soundscapes.

Studies on a Mobile Acoustic Emission Sensor Verification Device

Despite the fact that there are two ASTM standards, ASTM 1106 for primary calibration and ASTM E 1781 for secondary calibration of acoustic emission sensors, there is no company or institution currently offering calibration services based on these standards due to lack of availability or practical aspects. Another option is CEN ISO/TR 13115, which needs a large volumetric propagation medium leading to similar practical limitations as for the ASTM standards. Many acoustic emission users try to fill the gap of calibration services by using simplified verification standards. However, there is currently no verification standard suitable for using in the field. In that sense, one ambition of the project CalibrAEte is to establish a new standard for primary and secondary calibration that can be implemented in laboratories without a huge invest. Another objective of the project is to establish a verification procedure using a device to verify acoustic emission sensors by a wave-based approach in the field. The proposed approach uses a durable propagation medium that carries the elastic waves from a repeatable source to the sensor under test. This research focuses on such a mobile verification device in the sense that an optimal configuration is proposed. This contribution will present the major aspects and details of an acoustic emission sensor verification setup that can be used not just in the laboratory environment but also in harsh field conditions. It will be shown the performance of the reference sensor developed for this purpose along with the setup itself, which was partially developed by aid of numerical methods. Also, experimental results in different temperature conditions, will be presented to assess the robustness of the setup. It also be presented a sensitivity analysis of the setup using different sensors in different stages of degradation.

Research of Acoustic Emission Characteristics and Applicability of Artificial Intelligence for Source Localization

Within this research, the deployment of artificial intelligence for the localization of acoustic emissions (AE) is being investigated. In detail, series of experiments were conducted on a plate of granite with homogeneous and isotropic material behavior. Acoustic emissions could be generated by the breaking of pencil leads, as well as the impact of rice grains and steel balls on 81 specific positions with a distance of 50 mm in both horizontal axes to each other. The response of the system was recorded using 16 acoustic emission sensors, of which 8 sensors were positioned on each surface side of the plate. In further steps the data could be processed, so that the resulting characteristic signal features could be translated into spec-trograms and furthermore being used as input for the artificial intelligence. Additionally, the corresponding source locations appeared as labels. Both, the signal features as well as the labels, could be implemented in the process of supervised learning using a convolutional neural network (CNN). The results of this network indicated a validation accuracy of over 99% as well as a low loss value by classifying the inserted features to the right labels. Thus demonstrates the excellent applicability of artificial intelligence for the source localization of acoustic emission (AE).

Acoustic emission for monitoring of fatigue damage in concrete elements of wind turbine towers

Wind energy has become an important player in the energy transition in Germany. Towers of onshore wind turbines are often designed as hybrid structures: the lower part is made of prestressed concrete whereas the upper part is made of steel tubes. The tall structures are permanently subjected to dynamic loads. A research project of BAM as part of the joint project WinConFat-Structure focusses on the development of techniques to monitor fatigue damage evolution in the concrete part. Results of a previous project show that a combination of ultrasonic and acoustic emission testing can give an indication for critical conditions near the end of the fatigue life of the concrete. In the ongoing project acoustic emission sensors have been installed at the base and at the transition piece between concrete and steel of a hybrid wind turbine tower. Beside of acoustic emission measurement the sensor spacing allows for measuring the concrete ultrasonic velocity along the circumference in both levels. Additional measurements like strain, temperature, inclination, or acceleration allow for comparison of environmental loads and change of acoustic properties of the concrete. The paper focusses on first acoustic measurements recorded since December 2023 in comparison to operating data of the wind turbine.

Rapid detection of depression by volatile organic compounds from exhalation

Depression is a pervasive and often undetected mental health condition, which poses significant challenges for early diagnosis due to its silent and subtle nature. To evaluate exhaled volatile organic compounds (VOCs) as non-invasive biomarkers for the detection of depression using a virtual surface acoustic wave sensors array (VSAW-SA). A total of 245 participants were recruited from the Hangzhou Community Health Service Center, including 38 individuals diagnosed with depression and 207 control subjects. Breath samples were collected from all participants and subjected to analysis using VSAW-SA. Univariate and multivariate analyses were employed to assess the relationship between VOCs and depression. The findings revealed that the responses of virtual sensor ID 14, 44, 59, and 176, which corresponded respectively to ethanol, trichloroethylene or isoleucine, octanoic acid or lysine, and an unidentified compound, were sensitive to depression. Taking into account potential confounders, these sensor responses were utilized to calculate a depression detection indicator. It has a sensitivity of 81.6% and a specificity of 81.6%, with an area under the curve of 0.870 (95% CI = 0.816-0.923). Conclusions: exhaled VOCs as non-invasive biomarkers of depression could be detected by a VSAW-SA. Large-scale cohort studies should be conducted to confirm the potential ability of the VSAW-SA to diagnose depression.

Optical fiber sensors for acoustic emission monitoring

Acoustic emission sensors are most of the time based on piezoelectric sensors, which may not be suited for harsh environments or be too bulky to integrate a large number of them. We present here a system based on Fiber Bragg gratings on optical fibers to measure acoustic emission events. The use of such sensors enables a low intrusivity and measurements in extreme environments. The system is designed to enable measurements under varying environmental conditions, enabling its use in real applications.

Optimising Sensor Placement for Tool Condition Monitoring: A Comparative Analysis of Acoustic Emission Data

In the monitoring of machining processes and in the context of tool condition monitoring, the position of acoustic emission sensors affects the quality of the resulting information. This drives this comparative study which presents a comparison of acoustic emission data recorded simultaneously by a sensor attached to the side of the milling table and another sensor near the spindle during a tool wear test on a CNC milling machine. The signal information is clearly influenced by the different positions. Whereas the signals near the spindle are dominated by vibrations of the spindle motor during machining, these strong contributions do not occur in the recorded data of the machining process detected by the sensor mounted to the table. However, the position of the sensor near the spindle has the advantage of maintaining a constant distance to the source of the acoustic emissions of the machining process, as the sensor moves with the tool. In contrast, the sensor attached to the milling table is affected from the fact that the distance to the milling tool is constantly changing. This results in different dominant frequencies, that are analysed in this study. To analyse both the entire frequency range and the higher frequencies, two distinct approaches are employed, one without and one with the use of a high-pass filter. Furthermore, the recorded signals were used to calculate the characteristic features, namely the Partial Powers, thereby enabling assessment of the condition of the tool. The results demonstrate that this single feature type is sufficient for training a machine learning model to predict the tool wear, focusing on the flank wear width. This study analyses the efficiency of wear prediction with respect to both sensors and frequency spectra, and compares the results. In this context, the most accurate predictions were achieved with Gaussian process regression models.

Sediment Dynamics in the Energetic Nearshore Zone: Acoustic Remote Sensing and Model Validation

AbstractAcoustic backscatter data were collected in the shallow nearshore environment under breaking and non‐breaking waves, using a multi‐frequency sonar at 0.5, 1.0, and 2.0 MHz. The data are used to develop and test an acoustic inverse model for measuring suspended sediment concentration (SSC) in the presence of bubbles generated by breaking waves. The model leverages the contrasting frequency dependence of acoustic scattering by bubbles and sand, to simultaneously estimate SSC and bubble void fraction. Validation against in situ sediment measurements shows the acoustic technique can recover sediment concentration with good accuracy in both breaking and non‐breaking waves, unlike existing algorithms which only perform well in non‐breaking waves. Finally, data from the experiment are used to validate existing theories for suspended sediment dynamics under breaking waves, for which few previous data sets exist.

A novel optical acoustic fabric based on intensity-modulated flexible air-filled fiber adapter for sound sensing

This study proposes and demonstrates an optical acoustic fabric design method to detect sound, which is based on intensity-modulated flexible air-filled fiber adapter (FAFFA). A FAFFA was woven into the weft center of the polyamide fabric as weft, and the FAFFA was realized by connecting polymer optical fibers at each end of an flexible tube. The principle of the optical acoustic fabric is to allow the flexible tube to deform with the resonance of the polyamide fabric, subsequently changing the light intensity in the FAFFA. The macro-scale finite element method and ray tracing optical simulation investigated the operating principle of optical acoustic fabric and performance of FAFFA structures, and the influence of fabric parameters was evaluated by meso-scale finite element simulation. The yarn of the optical fiber acoustic fabric has low Young’s modulus and high flexibility. The proposed FAFFA structure is simple to prepare and has good maneuverability, which breaks away from the constraints of the traditional optical acoustic sensor membrane structure. This proposed optical acoustic fabric is easy to prepare, simple in principle and highly flexible. Along with wireless devices, the use of the optical acoustic fabric can be further extended to “hearing aids” for special individuals in wearable applications.

Localization of partial electrical discharges using compressive spherical frequency-difference beamforming.

Accurate localization of partial electrical discharges is essential for the diagnosis of high-voltage systems. The current study achieves this by employing an acoustic sensor array and a beamforming approach. The occurrence of a partial discharge is accompanied by the emission of high-frequency sounds in the ultrasonic range, making localization a challenging task requiring many sensors to avoid spatial aliasing. Compressive frequency-difference beamforming, as previously proposed, can be effective in addressing this issue. We expand the method to include near-field localization by utilizing a spherical wave and propose a two-step normalization process. This eliminates the bias associated with nonplanar waves and standardizes the field variables, thereby preserving only the phase and relative amplitude information. A distributed algorithm based on the alternating direction multiplier method is used to solve the associated convex optimization problem. The proposed method is demonstrated using simulated and experimental data.

High-resolution observations of shallow-water acoustic propagation with distributed acoustic sensing.

Distributed acoustic sensing (DAS), converting fiber-optic cables into dense acoustic sensors, is a promising technology that offers a cost-effective and scalable solution for long-term, high-resolution studies in ocean acoustics. In this paper, the telecommunication cable of Martha's Vineyard Coastal Observatory (MVCO) is used to explore the feasibility of cable localization and shallow-water sound propagation with a mobile acoustic source. The MVCO DAS array records coherent, high-quality acoustic signals in the frequency band of 105-160 Hz, and a two-step inversion method is used to improve the location accuracy of DAS channels, reducing the location uncertainty to ∼2 m. The DAS array with refined channel positions enables the high-resolution observation of acoustic modal interference. Numerical simulations that reproduce the observed interference pattern suggest a compressional speed of 1750 m/s in the sediment, which is consistent with previous in situ geoacoustic measurements. These findings demonstrate the long-term potential of DAS for high-resolution ocean acoustic studies.

ADRP-DQL: An adaptive distributed routing protocol for underwater acoustic sensor networks using deep Q-learning

Underwater Wireless Sensor Networks (UWSNs) face unique constraints due to their unstructured and dynamic underwater environment. Data gathering from these networks is crucial as energy resources are limited. In this regard, efficient routing protocols are needed to optimize energy consumption, increase the network lifetime, and enhance data delivery in these networks. In this work, we develop an Adaptive Distributed Routing Protocol for UWSNs using Deep Q-Learning (ADRP-DQL). This protocol employs the ability of reinforcement learning to dynamically learn the best routing decisions based on the network’s state and action-value estimates. It allows nodes to make intelligent routing decisions, considering residual energy, depth and node degree. A Deep Q-Network (DQN) is employed as the function approximator to estimate action values and choose the optimal routing decisions. The DQN is trained using off-policy and on-policy strategies and the neural network model. Simulation results demonstrate that ADRP-DQL performs well regarding energy efficiency (EE), data delivery ratio, and network lifetime. The results highlight the proposed protocol’s effectiveness and adaptability to UWSNs. The ADRP-DQL protocol contributes to intelligent routing for UWSNs, offering a promising approach to enhance performance and optimize energy utilization in these demanding environments.

Strategic deployment in the deep: Principled underwater sensor placement optimization with three-dimensional acoustic map.

Underwater acoustic sensors are vital for monitoring marine environments and detecting targets, but their optimal placement presents challenges, particularly in deep-sea environments. This paper addresses the question of determining the optimal sensor placement in a specific ocean region through a principled optimization approach. While previous studies mainly utilized heuristic algorithms without exploiting problem-specific structures, this work explores leveraging the complex three-dimensional acoustic environment through principled modeling and tailored optimization. Specifically, intricate three-dimensional multi-directional acoustic maps are constructed for each sensor. Based on these maps, the sensor placement problem is then cast as an integer linear programming, allowing the study to leverage established theoretical results from operations research. Additionally, an alternative algorithm with its performance indicator is presented to find near-optimal solutions efficiently and can empirically reach over 99% coverage of the optimal solution. Experimental results using real-life data from the South China Sea demonstrate the effectiveness of the proposed approach in achieving much larger detection coverage compared to random and empirical strategies. Notably, the alternative fast algorithm approaches the optimal solution in significantly less time. Furthermore, experiments show that any further simplification of this approach leads to the performance degradation.

Acoustic emission during the freezing of water or the melting of ice in aircraft

It is a frequently overlooked phenomenon that matter can produce sound emissions when undergoing a phase transition. The exact cause of this sound is until now not modeled and understood in depth, but it appears certain that sudden volume changes, cracks, friction, and other sources that occur during e.g. crystal decay/growth in melting/freezing processes must be considered. However, a generally accepted theory of the "sound of phase transitions" is lacking. In this paper we present a general introduction in this field. Besides the more fundamental aspects, the AE generation during the melting of ice can be used to detect the presence of this respective substance in technical structures. An interesting application is to detect the melting ice in aircraft. Adverse water accumulation, resulting from contamination and condensation, is regularly found in fuel tanks, and scheduled drainage procedures of this water after the melting of the ice are important to ensure the safety of systems and structures. A practical problem for drainage after flight is determining the right time to start the process. Beginning too early would mean that the remaining ice cannot be removed, and waiting too long presents an economic problem as it increases the aircraft's downtime. There is currently no technology that reports when all water has melted. To realize this goal acoustic sensors were attached to the skin of the tank walls, and the acoustic signals were intense enough to propagate through the aluminum sheets and coating and further analyzed. In this way, the completion of the melting of ice was determined by the time at which the acoustic emission stopped. We present measurements on a laboratory scale, the results from a climate chamber on relevant replica’s, and the outcome of a first campaign on an operational aircraft (Airbus A330).