In additive manufacturing technologies, the use of pastes and inks based on materials such as clay to create three-dimensional objects layer by layer has opened new possibilities in fields such as engineering and biomedicine. This review article aims to provide a comprehensive understanding of 3D printing of ceramic pastes through Direct Ink Writing (DIW), also referred to as Robocasting. DIW offers specific advantages for ceramic 3D printing, including the ability to extrude highly loaded pastes with customized rheological properties to accommodate a broad spectrum of ceramic compositions, varying from conventional clays to advanced ceramics. It is characterized by filament deposition control, which facilitates the fabrication of complex, porous, or customized architectures while simultaneously minimizing material waste. Through a bibliometric analysis of the literature published between 2020 and 2024, the most relevant studies regarding printing system architectures, ceramic paste formulations, and adjustment of parameters to obtain high-quality parts were identified. This work presents relevant and accurate explanations of the DIW technology, supporting researchers and industry professionals seeking to initiate or improve ceramic 3D printing processes for a wide range of applications.

While hardware selection is widely recognized as a key factor in machine learning performance, systematic and reproducible evaluation across heterogeneous and accessible platforms remains limited, particularly when jointly considering execution time, energy consumption, stability, and cost-efficiency. This work presents a unified and fully reproducible benchmarking framework for supervised learning, designed to enable controlled and comparable evaluation across diverse hardware environments. The proposed methodology enforces consistent training pipelines, fixed hyperparameter configurations, and repeated executions to ensure statistical reliability, while incorporating performance metrics such as execution time, power consumption, and energy usage, as well as performance-per-dollar. The framework is validated on a representative set of platforms, including CUDA-enabled GPUs, Apple Silicon (CPU/GPU), x86 processors, ARM-based embedded systems, and cloud-based environments, using convolutional, recurrent (RNN, LSTM, BiLSTM), and tree-based (XGBoost) models. The results reveal that hardware efficiency is strongly model-dependent. GPUs provide the highest computational performance and stability for parallel workloads, whereas Apple Silicon achieves superior energy efficiency with competitive execution times, particularly for recurrent architectures. The batch size analysis shows that performance can vary significantly depending on workload configuration, especially on CPU-based platforms, while epoch-based evaluation confirms that the measured performance reflects steady-state behavior rather than initialization overhead. In contrast, conventional CPUs and embedded systems exhibit significant scalability limitations for deep learning training, although they remain competitive for tree-based methods such as XGBoost, which demonstrates near hardware-independent predictive performance. These findings highlight the limitations of generalized hardware selection criteria and emphasize the need for model-aware and hardware-aware benchmarking. The proposed framework offers a practical and extensible foundation for reproducible, hardware-aware evaluation of machine learning systems, supporting informed decision-making in research, deployment, and cost-constrained scenarios.

Glass additive manufacturing (AM) is a developing technology, particularly in comparison to metals and polymers, both of which have had their processes and applications extensively studied. Its potential lies in fabricating complex, even micrometric, geometries that are difficult or impossible to achieve via traditional molding, as well as in producing components with unique optical properties. The diversity of AM techniques, alongside the challenges associated with the high melting point, rheological control, and fragility of glass, necessitates a comprehensive analysis of current developments. Accordingly, this review presents a systematic review, conducted in accordance with the PRISMA protocol, of recent literature regarding AM technologies that fabricate glass via particle fusion to form solid components. This review explicitly excludes techniques utilizing glass fibers as reinforcement, as that constitutes a separate field of inquiry. The results demonstrate sustained growth within the field, with a predominance of technologies based on photopolymerization and ink extrusion, both of which offer high resolution and microstructural control. Ultimately, this review establishes the current state of the art, identifying critical challenges and emerging lines of research to guide future development. It is intended to serve as a foundational reference for researchers and professionals seeking to initiate or expand their work in glass AM.

Este libro presenta tres alternativas para el análisis cinemático de mecanismos planos: método gráfico, análisis vectorial y análisis por el método de números complejos, con el fin de desarrollar una comprensión amplia y básica de la cinemática de los mecanismos. Está dirigido a profesionales y estudiantes de ingeniería como un referente que indica, a través de diferentes ejercicios desarrollados, alternativas para el análisis de mecanismos.En el primer capítulo y en la primera parte de los demás, se hace un breve análisis de los fundamentos necesarios para abordar las tres metodologías de análisis. Además, en la parte final del primer capítulo se hace una presentación de los mecanismos analizados cinemáticamente en los capítulos subsiguientes.Con el método gráfico del segundo capítulo se hace una presentación básica para el análisis cinemático de mecanismos en una posición particular. Este método además se utiliza como referencia para validar los resultados de otras metodologías como la vectorial o de número complejo presentadas en los capítulos 3 y 4.El análisis vectorial del capítulo 3 hace uso de vectores para representar las posiciones, las velocidades y las aceleraciones, así como para representar relaciones de dirección y sentido entre distintos elementos del mecanismo. De forma similar al capítulo 3, en el capítulo 4 se emplean números complejos para representar las magnitudes vectoriales, simplificando el cálculo de las cantidades involucradas.Para finalizar, en el capítulo 5 se presenta el uso de las herramientas computacionales, MATLAB o GNU Octave, para programar los modelos matemáticos desarrollados en los capítulos 3 y 4, y establecer el comportamiento de un mecanismo en función de las condiciones de entrada. Los resultados obtenidos en este capítulo son validados mediante el uso del método gráfico del capítulo 2.En general, este texto sobre cinemática de mecanismos planos proporciona una base teórica y práctica para comprender cómo se mueven y posicionan los componentes de un mecanismo, utilizando herramientas analíticas y gráficas para el estudio de sistemas mecánicos.

This work proposes and evaluates a supervised classification framework based on multichannel vibration signals to analyze two systems associated with wind turbines: rotor imbalance in a laboratory-adapted Enair E30Pro wind turbine and structural damage in a jacket-type structure. The methodology compares single-task and multitask learning models using spectral representations obtained from the Fast Fourier Transform (FFT) in log-magnitude format, the Short-Time Fourier Transform (STFT) in log-power format, and the Welch power spectral density (Welch PSD) in log-power format. During the single-task stage, TabNet (Attentive Interpretable Tabular Learning) models, a deep learning architecture designed for tabular data, were independently trained for each system and representation to establish baseline performance levels. Subsequently, a multitask scheme composed of a shared encoder and two task-specific classification heads was implemented: one for the jacket-type structure and another for the rotor. In addition to the original TabNet baselines, a capacity-matched single-task MLP control was evaluated using the same encoder architecture, preprocessing, optimization settings, epoch budget, and cross-validation folds as the multitask model. Relative to the TabNet baselines, the multitask model achieved higher performance in several configurations, particularly for the more challenging rotor task; however, the capacity-matched control showed that these gains cannot be attributed uniquely to multitask parameter sharing because the multitask model did not consistently outperform an equivalently structured single-task network. For the jacket-type structure, all three spectral representations achieved near-perfect classification performance, with Macro-F1 values above 0.99 in both the single-task and multitask settings. For the rotor imbalance problem, the multitask model using Welch PSD log-power achieved a global accuracy of 0.9125 and a Macro-F1 of 0.8909. Negative-control experiments further showed that replacing the auxiliary-task labels with random permutations reduced target-task performance in several configurations, indicating that informative auxiliary supervision influences the shared optimization process, although this effect did not translate into a consistent predictive advantage over the capacity-matched single-task control. The multitask architecture nevertheless provides a parameter-efficient joint formulation: one shared multitask network contains 4,481,354 trainable parameters, compared with 8,958,218 parameters required by two independent capacity-matched single-task networks, corresponding to an approximately 49.97% reduction in total model parameters. Overall, the results support multitask learning as a compact joint modeling strategy for physically different vibration-based classification problems, while indicating that its principal benefit in the present study is architectural efficiency rather than a consistently superior predictive effect attributable to positive cross-task transfer.