Jon García-Barruetabeña

Este artículo presenta una metodología de identificación de la anisotropía introducida por el proceso de fabricación mediante ensayos dinámicos. Se fundamenta en la identificación del comportamiento no lineal en puntos críticos de componentes de automoción. Una ventaja de esta metodología es que es especialmente útil para componentes manufacturados, donde queremos considerar el impacto del proceso de fabricación, especialmente en puntos críticos del componente, como cambios de secciones y curvaturas. Como resultado, podemos identificar cambios en las propiedades dinámicas derivado de la anisotropía en componentes de automoción clave para garantizar el confort de los usuarios, así como el comportamiento dinámico del vehículo.

The industrial sector has undergone significant digital transformation, driven by advancements in technology and the Internet of Things (IoT). These developments have facilitated the collection of vast quantities of data, which, in turn, pose significant challenges for real-time data processing. This study seeks to validate the efficacy and accuracy of edge computing models designed to represent subprocesses within industrial environments and to compare their performance with that of traditional cloud computing models. By processing data locally at the point of collection, edge computing models provide substantial benefits in minimizing latency and enhancing processing efficiency, which are crucial for real-time decision-making in industrial operations. This research demonstrates that models derived from distinct subprocesses yield superior accuracy compared to comprehensive models encompassing multiple subprocesses. The findings indicate that an increase in data volume does not necessarily translate to improved model performance, particularly in datasets that capture data from production processes, as combining independent process data can introduce extraneous ‘noise’. By subdividing datasets into smaller, specialized edge models, this study offers a viable approach to mitigating the latency challenges inherent in cloud computing, thereby enhancing real-time data processing capabilities, scalability, and adaptability for modern industrial applications.

This paper discusses the increasing amount of data handled by companies and the need to use Big Data and Data Analytics to extract value from this data. However, due to the large amount of data collected, challenges related to the computational capacity of machines often arise when performing this analysis to acquire relevant information for the organization, especially when we are using edge computing. The paper aims to train machine learning models using compressed data, with two compression techniques applied to the original data. The results show that models trained with compressed data achieved similar accuracy to those trained with uncompressed data, and different compression techniques were compared. The research extended a previous study by analyzing the use of autoencoders for compression and reducing both instances and dimensionality of the dataset. The accuracy rate of the models when trained with compressed data instead of original data was maintained.

Actualmente la mayoría de los datos que se recogen en las empresas y en los entornos de fabricación industrial a través de dispositivos IoT se procesan en la nube [1]. Dado el gran volumen de datos que cada empresa gestiona debido a la aparición del IoT, la computación en la nube no es la mejor opción para ciertos sectores como es el de la automoción [3]. Dentro de este sector, la calidad que el cliente final percibe está directamente relacionada con la cadena de montaje. Estas cadenas de montaje recogen una cantidad elevada de variables (temperaturas, presiones, bombas, etc.) y una predicción en tiempo real mediante pequeños gemelos digitales en este proceso evitaría costes tanto de material como de mano de obra. Hoy en día, realizar esa anticipación mediante modelos de inteligencia artificial en tiempo real es inviable debido a la latencia existente con el procesamiento en la nube. Es por eso, que existe una necesidad inminente de desarrollar aplicaciones que se desplieguen en el borde o 'Edge' de la red para el proceso de fabricación y pintado del automóvil que permitan la generación de gemelos digitales. En esta línea, la propuesta del concepto 'Edge Computing' [2] busca paliar en parte esta situación. Por este motivo, este artículo se ha centrado no solo en demostrar que al dividir los datos de un modelo grande generando pequeños modelos ‘Edge’ se mantiene la capacidad de predicción de cada algoritmo, sino que también se han realizado pruebas de estrés (límites, restricciones, etc.) en la tecnología hardware 'Edge Computing' disponible actualmente en el mercado que permite generar y procesar este tipo de modelos. Palabras Clave: Computación en el borde, Internet de las cosas, Inteligencia artificial, Industria 4.0 Currently most of the data collected in companies and industrial manufacturing environments through IoT devices is processed in the cloud [1]. Given the large volume of data that each company manages due to the emergence of IoT, cloud computing is not the best option for certain sectors such as automotive [3]. Within this sector, the quality perceived by the end customer is closely linked to the assembly line. These assembly lines collect a high number of variables (temperatures, pressures, pumps, etc.) and a real-time prediction by means of small digital twins in this process would avoid both material and labor costs. Today, to perform such anticipation through artificial intelligence models in real time is unfeasible due to the latency that exists with cloud processing. Therefore, there is an imminent need to develop applications that are deployed at the edge of the network for the automotive manufacturing and painting process that enable the generation of digital twins. In this regard, the proposal of the 'Edge Computing' concept [2] seeks to mitigate in part this situation. For this reason, this paper has focused not only on demonstrating that splitting the data of a large model by generating small 'Edge' models maintains the predictive capacity of each algorithm, but also stress tests (limits, constraints, etc.) have been performed on the 'Edge Computing' hardware technology currently available on the market that allows generating and processing this type of models. Key Words: Edge Computing, Internet of things, Machine Learning, Industry 4.0

This paper presents a robust method to estimate polymers’ damping, based on modal identification methods on frequency functions. The proposed method presents great advantages compared to other traditional methods such as the HPB method for polymeric materials where high damping or noise levels can limit their use. Specifically, this new method is applied on an experimental transmissibility function measured in a composite cantilever beam and the complex modulus is determined as a function of frequency. From this, a regenerated function is obtained based on the Euler–Bernoulli beam theory, and it is compared with experimental data. It can be concluded that the best way to apply the curve-fitting method for further testing of polymeric materials is when it is used with the whole frequency range by means of the MDOF method considering the residuals. In addition, this has the added advantage that the number of experimental tests to be carried out is much lower compared to using the SDOF method.

Recently, vibration control has been useful in various engineering fields such as aerospace, adaptive dynamic vibration absorbers, and infrastructure. Magnetorheological elastomer (MRE) is an interesting material for controlling and suppressing undesirable vibrations through the application of a magnetic field. The present study aims at analyzing the pre-structure of the magnetorheological viscoelastic core in the dynamic response of an MRE-sandwich structure. The forced vibration tests were performed under a non-homogenous magnetic field to evaluate the dynamic properties of the MRE-sandwich structure in a frequency bandwidth range of 0–250 Hz. Experimental results show that the proposed MRE-sandwich structures are capable of eliminating unwanted resonances due to induced magnetic field intensity in the activated region, especially at the fundamental mode. Moreover, results highlight that an oriented pre-structure in an MRE-sandwich has an attenuation effect on vibrations in the low frequency range. Additionally, the external magnetic field increased the structural vibrations damping capability by approximately 200%. In addition, the oriented pre-structures of the MRE core were also used to dissipate vibration. Consequently, they could potentially be used in vibration attenuation applications such as stop operations in dynamic structures.

Magnetorheological elastomers (MREs) have been developed as a new class of smart materials for such diverse applications as adaptive vibration control of systems and energy dissipation. The linear viscoelastic (LVE) behavior of MREs is expected to play a significant role in determining filler morphology (structure, size, and aspect ratio) and working conditions. The LVE behavior of pre-structures, especially anisotropic structures, has not been clearly explained. To this end, this study aimed to analyze the LVE behavior and dynamic properties according to different pre-structures of magnetic particles. This is based on MRE pre-structures oriented at 0°, 30°, 45°, and 90° using oscillatory shear excitation as a function of frequency and amplitude sweep tests. For the latter, the excitation frequency and magnetic field intensity were analyzed. These analyses were performed for both storage modulus and loss factor corresponding to 10% strain amplitude from the roughly level linear line of the LVE region. The results showed that increasing the frequency or magnetic field intensity to increase anisotropy reduced the LVE limit. For every oriented pre-structure and analyzed frequency tested, the magnetic field decreased the LVE limit. Thus, the orientation of the carbonyl iron particles in the matrix enhanced the viscoelastic properties.