Browse Topic: Design Engineering and Styling

Items (46,222)
Achieving best-in-class Noise, Vibration, and Harshness (NVH) in electric powertrains demands a paradigm shift in development methodology. This paper presents a practice-oriented overview of simulation methods in NVH development methodology for electric drive units. This includes target cascading and multi-objective optimisation, and by attacking NVH at the source using KPIs early in the design cycle, significant reductions in development time and reliance on traditional testbed loops are realised. Machine learning (Neural Network) algorithms are utilized to find the best-in-class design, using multi-objective optimisation as well as refining simulation accuracy by adding tolerance effects while target cascading ensures alignment of system-level performance objectives down to subsystem contributions. Combined, these strategies enable rapid and robust NVH optimisation, using simulation for next-generation electric powertrain development. Several applications and real-life examples
Mehrgou, MehdiGarcia de Madinabeitia, InigoGraf, BernhardGojo, Josef
Simulations can only be searched, reused and leveraged as training data for machine learning methods if suitable metadata are related. Manually obtaining these metadata is time-consuming and requires expert knowledge. Consequently, there often is a lack of metadata and this prohibits the reutilization of simulation data. Therefore, automated frameworks for metadata extraction are essential to obtain metadata information quickly, effortlessly and cost-efficiently. At present, there are no toolboxes for Finite-Element-Simulation data. Nevertheless, machine learning methods are a promising solution for this task. Training classical supervised machine learning methods for metadata generation often faces the lack of labeled data since manual labelling can be very costly. Therefore, rule-based extraction algorithms are used as an alternative for fundamental metadata extraction. For more enhanced tasks they are often not feasible. Active Learning is a suitable technique to overcome this
Luegmair, MarinusGröttrup, Sören
The simulation of structure-borne energy flow within a full vehicle trimmed body at mid and high frequencies has always been a challenge due to the large computational cost associated with standard deterministic simulations. This is a particularly pressing problem given that the electrification of the vehicles is extending the presence of structure-borne sources to higher frequencies. While the improvement of computational hardware has allowed OEMs to shift the limit of standard Finite Element (FE) approaches to higher frequencies, no methods have been proposed in the literature that tackle the full frequency range for industrial-sized problems. In this paper, a simulation methodology that uses wave-based processing of the original low-frequency finite element input deck to compute the coupling loss factors is proposed to model structure-borne noise in complex systems at mid and high frequencies. The methodology is validated against numerical and experimental data.
Errico, FabrizioLegault, JulienMordillat, PhilippeZerrad, Mehdi
Monitoring inputs and states of a structural dynamic system is often challenging, as direct measurements are costly or even infeasible. A virtual sensing methodology is presented for jointly estimating the input and state of a structure when subjected to multi-directional base excitations. The approach uses a tuned Kalman Filter combined with a model-order reduction of the system model to ensure a low computational cost whilst allowing accurate estimation from a limited number of acceleration measurements. This enables real-time virtual health monitoring strategies and reduction in instrumentation during data acquisition without additional information such as location and direction of application about the inputs. The proposed methodology is validated numerically and experimentally using a notched aluminum beam excited on a multi-directional shaker table, driven simultaneously in two in-plane directions. The study demonstrates accurate full-field estimation of multiple responses along
Salazar Colunga, RodrigoPandiya, NimishDindorf, ChristianNaets, Frank
When developing a vehicle, the overall body stiffness is an important parameter to be estimated for several automotive attributes. As a complement to the traditional experimental and computational static torsional stiffness assessment, an improved method has been developed to evaluate the body stiffness when driving the vehicle on a test track. This method, valid for both test and simulation, is called Opening Distortion Fingerprint (ODF) and uses the so-called Multi Stethoscope (MSS) to measure the dynamic distortion in each body closure opening and cross section. For evaluating the distortion, from both test and Multi Body Dynamics (MBD) simulation data, the Evaluation-line (E-line) method is used. The E-line method is a linear approach. Consequently, it is only valid in the absence of large rigid body rotations of the vehicle body. Therefore, to assess the validity of the ODF method, it is crucial to identify the frequency at which the distortion results become invalid due to rigid
Olger, EmmaLindkvist, LisaPiiroinen, PetriKarypidis, JohnPena, MiltonBäcklund, JesperAppelgren, PeterMarberg, HenrikUgale, PravinWeber, Jens
This paper presents an analytical model for three-phase Permanent Magnet Synchronous Motors (PMSMs) based on Magnetic Equivalent Circuits (MECs). The approach combines a reduced magnetic network, formulated in the complex domain to simplify the mathematical development, with an offline parameter estimation procedure systematically applied for different harmonic orders. This enables the model to capture the spatial dependence of permeance variations and reproduce inductance and magnetic flux nonlinearities, while maintaining generality, physical interpretability, and computational efficiency. Numerical simulations are compared with Finite Element (FE) results to validate the model’s ability to predict current and torque harmonics and the resulting radial electromagnetic forces, demonstrating its suitability for fast Noise, Vibration, and Harshness (NVH) analysis and vibroacoustic optimization.
Luciano, LudovicaDoria-Cerezo, ArnauSalamone, Nicolò
Space vector pulse width modulation (SVPWM) induces common-mode voltage (CMV) in three-phase voltage-source inverters, producing steep voltage edges that can lead to high leakage currents. In electric drive applications, these currents accelerate motor bearing degradation and may cause winding insulation failure. Active-zero-state PWM (AZSPWM) and near-state PWM (NSPWM) have been proposed as alternative modulation strategies to mitigate CMV and reduce drive degradation. This paper investigates the noise, vibration, and harshness performance of AZSPWM and NSPWM in comparison with conventional SVPWM. The proposed CMV reduction schemes are evaluated in terms of both CMV mitigation and their impact on high-frequency sideband vibration harmonics. Experimental results demonstrate that the CMV reduction strategies are highly effective in lowering CMV levels relative to SVPWM; however, this benefit is accompanied by an increase in vibration levels, which may adversely affect the mechanical
Khamis, Mahmoud AlyTatar, Andrei AlexandruRepecho, VictorDoria-Cerezo, Arnau
The vibro-acoustic performance of a vehicle is a critical factor in customer perception of quality and comfort, yet optimizing for Noise, Vibration, and Harshness (NVH)—specifically road noise—presents a persistent challenge in the modern automotive development cycle. While advanced Finite Element Method (FEM) analysis is essential, the increasing complexity and volume of CAE simulation data often overwhelm manual interpretation, potentially leading to prolonged development times or compromises in final comfort quality. To address these challenges, this paper introduces the application of CDH/ACE (Autonomous Computational Experiments), a framework that integrates conventional CAE simulation workflows with advanced machine learning in an iterative, cyclic process. This creates an exceptionally user-friendly and self-correcting system that autonomously defines, performs, and learns from computational experiments. By leveraging machine learning algorithms to build robust predictive models
Visser, Rene
Reconstruction of acoustic radiation from vibrating structures is central in vibroacoustics, as full-field sound information is essential for identifying radiation mechanisms and improving structural-acoustic performance. Conventional microphone-based measurements are limited by spatial sampling constraints and high experimental cost, while purely numerical approaches such as Finite Element Method (FEM) simulations offer flexibility but are strongly affected by parameter uncertainties, discretization errors, and imperfect boundary conditions. To overcome these drawbacks, this work develops a hybrid time-domain framework to reconstruct the radiated acoustic field by coupling vibration measurements to a FEM-based vibroacoustic model. The FEM model is reduced using Krylov subspace projection, yielding a compact state-space representation that captures the dominant vibroacoustic modes while remaining computationally efficient for sequential data assimilation. The acoustic radiation domain
Dong, LuyaoCai, YinshanDenayer, HervéDeckers, Elke
An application of a Non-Parametric Variability Modeling (NPVM), as introduced by Pr. Soize, of a full vehicle road noise simulation, is an opportunity to highlight some applicative issues of such a stochastic approach. First, the convergence of the stochastic computations is considered by introducing the probabilistic modal density of the considered model as an indicator of the system intrinsic dynamic behavior. Since the probabilistic model induces a spread of modal frequencies, the upper range shows a lack of modes, deviating from the actual system modal density. The study of this deviation leads to the modal truncation criterion required to achieve a relevant probabilistic modal density in a targeted frequency range. The required margin in order to achieve a proper convergence of the probabilistic problems appears larger than expected. Then, using appropriate parameters, road noise simulation is investigated in the framework of the stochastic modeling. After the capability of the
Gagliardini, LaurentGlandier, ChristianBauer, EricStraka, AndreasFiedler, Uwe
In the automotive industry, controlling noise transmission through vehicle components is essential for passenger comfort and regulatory compliance. Traditionally, Transmission Loss (TL) is estimated using simplified CAD-based metrics, which lack accuracy at high frequencies and for complex assemblies. Modeling complex vehicle components introduces challenges, such as representing fluid-structure and trim interactions, with spatially varying trim thicknesses. This study presents an industrial application implementing the Virtual SEA (Statistical Energy Analysis) method to evaluate TL for a firewall. The study discusses strategies for subsystem adaptation and analytical trim modeling, highlighting the importance of managing spatial averaging effects. The proposed workflow integrates laboratory measurements of trim materials, advanced subsystem definition, diffuse sound field (DSF) excitation and radiation in free-field condition. Virtual SEA results are systematically validated against
Orselli, JosephJacquemin, GaetanPark, MyeongMan
By using a fully trimmed vehicle body as flexible body, imported through a Modal Neutral File (MNF), in a complete vehicle Multibody Dynamics (MBD) analysis, the simulation setup gets considerably closer to the test conditions compared to only using a linear Finite Element Method (FEM) approach. Since the MBD analysis includes gravity, rigid body modes of the vehicle and the nonlinear behavior of the wheel suspension, it brings the correlation between simulation and test to a new and more comprehensive level. As correlation criteria, the results of the so-called Multi Stethoscope (MSS) are used. The MSS captures the time history of distortion in all body openings and cross sections and enables a detailed stiffness evaluation of the body using the so-called Opening Distortion Fingerprint (ODF). The ODF gives the quasi-static response while the Operational Deflection Shape (ODS), which is another result of the MSS measurements, reflects the dynamic response. Apart from the different
Lindkvist, LisaOlger, EmmaPiiroinen, PetriKarypidis, JohnPena, MiltonBäcklund, JesperAppelgren, PeterMarberg, HenrikUgale, PravinWeber, Jens
The virtual development of Electric Drive Modules (EDMs) for Battery Electric Vehicles (BEVs) requires proven and predictive methodologies. One part of the development investigates the vibro-acoustic assessment for the low- and high-frequency ranges within the targeted operating range. The efficient use of such a methodology requires an understanding of the accuracy and validity of the achievable results, as well as the derivation of suitable improvement measures for goals that have not been achieved. The use of reference data from experimental investigations and a detailed root cause analysis (RCA), to directly link a specific response and behavior to the excitations, modal content, and transfer functions, is an essential and non-trivial part of the methodology development. This paper describes the development of such a methodology using the example of a new EDM virtual model for Noise, Vibration and Harshness (NVH) analysis, including the simulation approach, validation, and
Klarin, BorislavPevec, DenisResch, ThomasEsposito, SaraD'Alessandro, VincenzoSpanu, Giorgio
Recent advancements in system-level NVH (Noise, Vibration, and Harshness) development methodologies have improved target cascading and enabled more efficient system-level optimization. Dynamic substructuring facilitates the virtual integration and modification of multiple subsystems and the prediction of changes in overall transfer functions. In practical automotive applications, advanced frequency-based substructuring has been applied to virtually modify system parameters, such as mass and stiffness, at multiple points in a target system, allowing prediction of the resulting effects and optimization of parameter changes without physical intervention. This study extends the methodology by introducing an enhanced substructuring approach capable of addressing not only basic parameter modifications but also large-scale structural changes. The proposed process involves identifying the characteristics of a base system assembly and a target subsystem, decoupling the subsystem from the
Cho, MunhwanBoelens, JelleReichart, Ronde Klerk, DennisAhn, Jiho
For analysing flow and acoustic induced structural vibration, a fully run time coupled framework combining a hybrid CFD-CAA approach with a modal response simulation was validated and presented at the ISVNH 2022 (SAE Technical Paper 2022-01-0938). In this paper i We apply this CFD–CAA–modal coupling method to a series-representative bonnet geometry and demonstrate its capability to capture flow and aeroacoustically driven vibration with two-way coupling. ii We analyse the modal properties of the bonnet and show that confined air volumes beneath the bonnet can introduce significant fluid loading effects, which are already embedded in experimentally validated FE modal models and must therefore be treated carefully in two-way coupled simulations. iii We validate the fully coupled aeroelastic simulation against wind-tunnel measurements with undisturbed inflow, show close agreement with the measured vibration response and analyse that the dominant excitation is in this case from below the
Schwertfirm, FlorianOcker, JoergHartmann, Michael
This work presents a modular engineering methodology (DiPhyBa - Digital Physical Balance) for the virtual validation of Noise, Vibration, and Harshness (NVH) performance in automotive development. The approach addresses the inefficiency of repeated physical testing across vehicle variants by introducing a structured two-phase process—Launcher and Reskin—centered on quantitative performance indicators with formal acceptance thresholds. In the Launcher phase, a digital replica of the base vehicle is built and iteratively correlated with physical test data. Validation is governed by objective indicators of confidence, conformity, and correlation, each evaluated against predefined thresholds. Once validated, the model becomes a certified reference, enabling its reuse across derivative configurations in the Reskin phase. Physical testing is only required if indicators fall below threshold, with a final gate test on pre-series vehicles ensuring industrial robustness. DiPhyBa formalizes the
Celiberti, LuciaCamia, Andrea
Achieving favorable Noise, Vibration, and Harshness (NVH) and durability performance in vehicles requires sufficient static and dynamic stiffness of the Body-in-White (BIW). Virtual development of BIW performance targets during the early design stages is essential to minimize costly modifications in later phases. In the automotive industry, full-scale finite element models are widely used for this purpose, offering high fidelity and enabling comprehensive performance evaluations. However, their complexity and high computational cost limit their practicality for early-stage sensitivity and optimization studies. Beam-based models offer a faster alternative; however, conventional beam formulations based on Euler–Bernoulli or Timoshenko beam theories often fail to capture the complex deformation behaviors of thin-walled structures, which are typical of BIW designs. This typically results in poor correlation with detailed models unless artificial joint flexibility is introduced at
Kim, Jin HongGang-Won, Jang
This study presents a high-fidelity NVH (Noise, Vibration, Harshness) analysis model development process for EV traction motors. The proposed process consists of two main components: Path advancement through structural stiffness tuning, and Source advancement, focused on the motor’s excitation mechanisms. Model accuracy was validated through comparison of simulation results with dyno experiment data, with particular focus on the 24th-order electromagnetic vibration observed in an 8-pole, 48-slot motor. Path advancement was achieved through modal correlation between experimental results and finite element (FE) analysis. Nine modal experiment and simulation stages were conducted, ranging from individual components to the complete motor assembly. Mode shapes were compared using the Modal Assurance Criterion (MAC), and natural frequencies were matched within a 5% error margin by adjusting FE material properties. For the 24th-order electromagnetic vibration, simulation results agreed with
Kim, DongheeKim, Dong-JunLee, SangHanKim, Seon HyeongHwang, Seung GyuValente, GiorgioParisouz, ShahriarHalse, Christopher
In recent years, the automotive industry has actively explored the application of various AI-based models such as Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) networks, Autoencoders, and Transformers to improve defect detection rates at the End-of-Line (EOL) stage. However, implementing these approaches in the Noise, Vibration, and Harshness (NVH) area face several practical challenges: ① extended evaluation times compared to other data types, which limit the quantity of training data and lead to overfitting; ② label imbalance caused by the relatively small amount of defect data; ③ reduced labeling accuracy due to human error; ④ decreased robustness under domain shifts such as changes in jig fixtures, test environments, and signal-to-noise ratio (SNR); ⑤ diminished model reliability when new defect arise during development; and ⑥ constraints imposed by compatibility requirements with existing test equipment. This study proposes a Convolutional Autoencoder (CAE
Park, Jun-SeoJo, Hyeon-ChoelCho, In-JeSeo, Jae-YongYoo, Seong-Sik
To minimize noise caused by interior components rubbing against each other, automotive materials are usually tested in advance with the established stick-slip method according to VDA standard 230-206. This procedure is widely used for soft materials, upholstery and plastics. However, it is limited to constant climatic and selected loading conditions. Contrary, in real application, changing climates and dynamic excitations can nevertheless trigger noise issues even in materials rated as suitable in the prior tests. To address this gap, a new test method has been developed that evaluates the stick-slip behavior of material combinations for a wide range of loading and climatic conditions. Conducted in a climate chamber with a standard stick-slip test bench, the procedure applies sinusoidal excitations, dynamic climatic shifts and advanced data analysis. In addition to the usual results the new method also evaluates realistic scenarios such as starting a vehicle in different seasons or
Fritz, SusanneStrangfeld, Martin
Noise phenomena in automobiles caused by the stick-slip effect are increasingly among the most frequent reasons for customer complaints and therefore represent a critical vehicle quality attribute. To proactively address such issues, stick-slip testing of contacting material pairs is commonly applied during development. However, the predictive capability of current stick-slip test methods remains limited, particularly when highly flexible materials and realistic, stochastic excitation conditions are involved. The flexibility of sealing systems often allows the actual relative motion at the contact interface to be accommodated through adhesion and elastic deformation, thereby delaying or even preventing sliding. To date, this effect has not been represented by any characteristic parameter in conventional stick-slip testing. Instead, existing evaluations focus exclusively on the analysis of occurring stick-slip oscillations. For the initiation of stick-slip phenomena, however, not only
Strangfeld, MartinFritz, SusanneWeber, JensRosell, Anneli
Gyroscopic effects split circumferential traveling-wave resonances of rotating structures into forward and backward branches. This work first analyzes the splitting in the co-rotating (Lagrangian) frame to provide physical intuition for the evolution of the two branches with spin speed. A transformation to the inertial (Eulerian) frame is then derived, showing that the observed frequencies are shifted by a kinematic Doppler-like term that acts with opposite sign on the forward and backward waves, leading to different Campbell-diagram slopes depending on the observation frame. The resulting framework is validated experimentally on a freely rotating, unloaded tire using two complementary sensing modalities: wireless on-tire accelerometers (co-rotating view) and a scanning laser Doppler vibrometer (inertial view). A frequency-domain SVD-based identification (FDD/ODS-SVD) is used to extract poles and deformation patterns over a range of spin speeds, enabling Campbell diagrams in both
del Fresno Zarza, JavierNaets, Frank
Electric vehicle subsystems, including powertrains, electric motors, and gearboxes, pose new challenges in achieving stringent acoustic performance targets for both interior and exterior noise. These challenges are intensified by increasingly demanding customer expectations regarding interior acoustic comfort, which encompasses the reduction of intrusive noise sources and the enhancement of overall sound quality across a broad frequency spectrum. A primary concern associated with electric vehicles subsystems is the generation of high-frequency tonal noise, commonly referred to as whine noise, which can significantly impact acoustic performance and passenger comfort. High-frequency whine noise propagates through multiple transmission paths and can be effectively attenuated at the source through encapsulation strategies, which also contribute to broadband noise reduction across a wide frequency spectrum. To predict the acoustic performance of encapsulation, a coupled simulation approach
Amichi, KamelCalloni, Massimiliano
Vehicle sound packages are usually designed to provide a given level of vehicle Noise, Vibration, and Harshness (NVH) comfort, within weight and cost constraints. Optimal comfort results can be obtained by considering the interaction of all the parts as a full physical system. So far, extensive research has already been performed and published on optimizing vehicle sound packages to achieve effective noise reduction at lowest cost and weight. Nowadays, due to the urgency of the transition to carbon neutrality, sound packages must also address the reduction of the full vehicle life cycle carbon emissions. Sound package components should use materials that have a low emission impact during production and that are suitable for recycling at the end of the vehicle’s life. This entails reconsidering the material solutions chosen for the sound package as a whole, rather than for each individual component. This article describes possible differentiations in the design of a sound package
Courtois, TheophaneCardillo, MarcoCriscione, MattiaGerges, YoussefMassocco, Andrea
Noise, Vibration, and Harshness (NVH) performance is critical in the automotive development process, yet identifying the true root causes of unwanted dynamic behavior remains a challenge in full vehicle or system-level finite element (FEM) models. This work demonstrates how Frequency Based Substructuring (FBS) provides an efficient framework for understanding NVH phenomena and facilitates new root cause analysis (RCA) types and processes. To begin, we prove the numerical accuracy of the FBS algorithm deployed in the presented investigation by comparing its results with those obtained with superelements and without substructuring. We point out that because the used FBS process starts with a modal representation of the components rather than their frequency response functions (FRF) a different class of RCA type becomes available. Then we introduce new RCA types starting with an analysis named Modal Influence (MI) that reveals the effect of the modes of any component on a certain response
Herbst, Markus
Part- or component-level tests are commonly performed by Tiers and OEMs to investigate the NVH behavior and loading mechanisms. However, because test bench dynamics differ from those of the actual vehicle environment, correlating measured sound, acceleration and forces between bench and vehicle often proves challenging. Blocked forces offer a way to address this issue, as they provide test bench and vehicle independent load representations. This effectively enables different Tiers to deliver consistent load data, which OEMs can then use to better tune excitation and noise transmission on their vehicles. This paper focuses on 2 test bench compensation techniques, involving pure test and a simulation models of the tire to obtain accurate blocked-forces. The compensation techniques are validated on four testbenches of different companies.
Reichart, Ronde Klerk, Dennis
Electric high voltage (HV) cables are commonly used in automotive applications and very prominently in electrified vehicles. These cables are potential flanking transmission paths for structure-borne sound in a broad frequency range and must therefore be included in the NVH design process. Electrical high voltage cables exhibit non-linear mechanical characteristics, when exposed to significant bending the internal geometry of the cable will change and a curvature dependent bending stiffness will result. The electrical cables envisaged in the current publication feature a helically wound stranded aluminium wire core. This conductive core is covered by, in sequence, a silicone rubber insulation, a braided aluminium wire shield with aluminium foil to minimize electromagnetic interference and a silicone rubber outer sheath. An extensive measurement campaign was carried out to dynamically characterize cable specimen of different lengths and cross sections in terms of multi-degree of freedom
Nijman, EugeneBuchegger, BlasiusBöhler, ElmarZeller, BernhardRejlek, JanFaksa, LukášLukavsky, David
Acoustic user interfaces and audio experiences are among the leading comfort factors in new vehicle interior designs. OEMs are more and more focusing on loudspeaker design and positioning, to provide the most immersive experience to the customers. The industrial target is to be able to predict the performance of an audio system in early design phases. This paper presents an integrated vibro-acoustic methodology enabling early-stage prediction of loudspeaker performance in real vehicle conditions. The approach combines electromechanical characterization, a hybrid loudspeaker calibrated model valid across the audible range and coupled FEM/BEM/SEA simulations to capture the loudspeaker response in the vehicle’s cabin considering door-installation effects and cabin acoustics. The method is validated experimentally on a rear-door loudspeaker installed in a production vehicle, showing strong correlation with measured SPL. A final application case demonstrates its capability to assess the
Zerrad, MehdiErrico, FabrizioMordillat, Philippe
Recent studies indicate that the door system plays a significant role in the interior noise levels of newly developed vehicles. This research investigates the noise transmission paths through the door system and identifies effective strategies for improvement through a combination of door buck testing and simulation. Specifically, in this study, the finite element method (FEM) was employed for door buck simulation, and the model was validated against vibration test results. Subsequently, acoustic analysis tools were utilized to correlate with noise testing, thereby establishing a process to ensure simulation accuracy. The sound insulation performance for the main areas of the door was experimentally evaluated, and a simulation model with good correlation to these test results was developed. By utilizing both experimental and simulation results, the principal transmission paths were identified, and appropriate improvement strategies for these paths were investigated. The validated
Chae, Ki-SangJang, JinungJeong, HojungDo, HyuncheolHan, JinwooYi, JaebokBak, Seong-JaeJeong, ChanHee
As acoustic requirements for NVH trim components become increasingly constrained by mass, cost, and sustainability targets, traditional approaches to inner dash design based on spatially averaged Transmission Loss (TL) metrics are reaching their practical limits. In fully built vehicles, the acoustic performance of the inner dash is governed by its global insulation capability but also by strong spatial heterogeneity and its interaction with spatially distributed noise sources such as the power unit, gearbox, and tyre-road excitation. This paper presents a test-based methodology for the spatial optimisation of inner dash acoustic performance using reciprocal holography. By applying a calibrated sound power source within the vehicle cabin and measuring the reciprocal response in the engine bay and wheel-arch regions, a high-resolution spatial Transmission Loss “hologram” of the inner dash is obtained under in-situ conditions. The resulting spatial data enables the identification of
Harry, EvanEandi, Giacomo
Noise pollution is a major environmental and health challenge, yet its strong spatial and temporal variability makes comprehensive mapping highly complex. Current approaches under the European Noise Directive (END) provide only partial coverage and often lack temporal dynamics. The NoiseSphere project, funded by the Austrian Research Promotion Agency FFG, develops an AI-based methodology for dynamic, large-scale noise prediction and mapping. A machine learning model is trained on heterogeneous data sources, including semantically enriched open Sentinel-2 satellite imagery, OpenStreetMap road data and existing noise maps. The model is refined through integration of noise emission data and validated using targeted in-situ measurements. A case study in an urban environment (Graz, Austria) demonstrates the model’s applicability. By combining remote sensing, traffic dynamics, and machine learning, NoiseSphere enables predictive noise mapping even in regions not covered by current
Girstmair, Josef
Items per page:
1 – 50 of 46222