Edwuin Carrasquero

Industrial machines, such as centrifugal pumps, accumulate large volumes of signals under healthy operating conditions long before representative fault data become available. However, the value of these normal-operation signals for subsequent fault diagnosis remains insufficiently explored. This work introduces a One-Class Signal World Model (OC-SWM), whose encoder is learned exclusively from healthy-operation data and subsequently transferred to diagnostic tasks when only a limited number of fault examples are available. The method constructs a sequence of latent states from each healthy vibration signal with a teacher–student–predictor architecture to learn their dynamics through latent-state prediction. The resulting encoder is then coupled with an attention-based readout and adapted using a small set of fault signals. Experiments on a ten-stage vertical centrifugal pump under six hydraulic regimes show that one-class pretraining yields a macro-F1 of 0.814 with only 30 labeled examples per health condition and a normalized area under the label-efficiency curve of 0.947, with the proposed method outperforming contrastive and adversarial alternatives and matching a regime-conditioned variant with lower dispersion across folds. Furthermore, the pretrained representation captures the structure of the hydraulic regimes before any fault-supervised adaptation, even though regime information is not explicitly provided during pretraining. This finding indicates that OC-SWM captures physically meaningful information about the machine operating state while providing a reusable representation for label-efficient fault diagnosis.

Phase space reconstruction (PSR) converts a vibration signal into a multidimensional trajectory encoding the nonlinear dynamics of rotating machinery, yet existing PSR-based deep learning approaches rasterize this trajectory into an image before classification, discarding its inherent temporal ordering. This paper proposes feeding two derivative-augmented PSR constructions—the Approximated Phase Space (APS) and the Classical Phase Space (CPS), directly as sequential inputs to a Long Short-Term Memory (LSTM) neural network, exploiting the structural compatibility between the column-sequential organization of the PSR matrix and LSTM’s temporal processing, and avoiding the image-conversion step used in CNN-based PSR approaches on the two datasets studied here. Each PSR matrix is augmented with the first and second derivatives computed via the Savitzky–Golay filter, extending the feature space to a derivative-augmented state representation that jointly encodes signal amplitude, rate of change, and curvature. Both constructions are validated through k-fold cross-validation on a 10-class spur gearbox dataset and a 7-class roller bearing dataset, both acquired under variable speed and load conditions. File-level classification via majority voting over feature pieces yields 99.7% accuracy for the gearbox and 99.7% for roller bearings using CPS features, and 94.6% and 99.0%, respectively, using APS features. CPS consistently outperforms APS by 9–14 percentage points at the piece level, attributable to its overlapping lag structure, while the majority voting substantially closes this gap at the file level. These results show that, on the two datasets evaluated, PSR matrices can be fed directly as sequential inputs to an LSTM without image conversion, yielding competitive fault classification accuracy, and that derivative-augmented phase space features provide a computationally efficient representation for fault diagnosis in rotating machinery.

Objective. To determine how bamboo loadings (2.5–5 wt%) and compatibilization with PBAT-g-MAH (BP-1, 10 wt%) affect melt flow and early-time mineralization of PLA biocomposites under near-ambient soil–compost conditions (ASTM D5988), while using PBAT-g-GMA (BP-2) only as a melt-flow screening reference. Methods. Melt flow index (MFI, ASTM D1238, 2.16 kg; 190/210/230 °C) was first measured for neat PLA and PLA/BP-1/BP-2 blends to select a printable matrix. PLA/10BP-1 composites containing 2.5–5 wt% bamboo were then compounded, extruded as bars for biodegradation tests, and validated by FFF printing. Biodegradation was quantified from titrimetric CO₂ evolution in soil–compost reactors at 21 ± 2 °C and pH ≈ 7 (triplicate specimens plus triplicate blanks; mean ± SD and endpoint statistics). ATR-FTIR was used to support mechanistic interpretation. Results. BP-1 markedly increased MFI relative to neat PLA, whereas BP-2 remained close to the neat matrix, consistent with epoxy-driven coupling that can raise viscosity. Under ambient burial, all materials exhibited very low mineralization over 0–23 days; PLA/10BP-1/2.5B and PLA/10BP-1/5B showed a slight increase in net CO₂ evolution compared with neat PLA, but the differences remained modest and within the experimental uncertainty, reflecting a balance between bamboo’s pro-hydrolytic effect and the sealing action of PBAT-g-MAH compatibilization. Significance. The data delineate a printing–degradation window in which PLA/10BP-1 with 2.5–5 wt% bamboo combines easy processing and short-term durability while preserving industrial compostability at end-of-life.

The present research aims to evaluate the levels of discrimination toward students with Autism Spectrum Disorder (ASD) at a private educational unit in Ecuador, as well as to raise awareness in the educational community about the specific abilities and needs of these students. To support the study, various learning theories were employed, such as Bandura's social learning theory and self-efficacy, Piaget's experiential learning, and the theory of mind proposed by Cohen, Leslie, and Frith. These provided a robust, multidimensional theoretical framework guiding method selection, addressing both processes of observation and imitation and the importance of self-efficacy and emotional interpretation in inclusive educational contexts. A mixed-methods approach was adopted, integrating qualitative and quantitative techniques to obtain a comprehensive, detailed view of the phenomenon. Data collection included semistructured interviews with two teachers specialized in basic education. Additionally, surveys were applied to 34 adolescents aged 14 and 15, of whom 3 have ASD. The study's results revealed that negative peer attitudes—shaped by lack of knowledge, stereotypes, scarcity of resources, and limited personalized support—constitute the most significant risk factors for school discrimination. It is concluded that immediate intervention strategies must be implemented through collaborative methodologies and teacher training programs to foster an inclusive and respectful school environment.

This study evaluates the competitiveness of small- and medium-sized manufacturing enterprises (PYMES) through a meta-analysis that explores the role of technology and quality. Through a comprehensive literature review, 24 eligible studies, selected after applying specific criteria in a systematic search of the Scopus database, were identified and analyzed. A random-effects model was used to combine the effect sizes of the selected studies, and the heterogeneity among them was assessed using widely accepted indicators such as I², H² and tau², which confirmed a moderate variability among the studies. The results of the meta-analysis indicated a pooled effect size of 0.53 (95% CI: [0.50, 0.55]), suggesting a significant positive relationship between innovation and competitiveness, as well as between quality and competitiveness. Quality was identified as the most relevant competitive priority, while the use of state-of-the-art technologies was highlighted as a significant risk factor in the context of digital transformation. In addition, topic analysis was performed using the latent Dirichlet allocation (LDA) model, implemented with the topicmodels package in RStudio 2024.04.2. To ensure accuracy of the analysis, the texts were preprocessed using cleaning and tokenization techniques, which included the removal of punctuation, numbers and empty words. This thematic analysis identified key patterns related to innovation management, operational strategies and integration of digital technologies. The themes generated revealed that manufacturing PYMES prioritize quality as a source of competitive advantage, facing significant challenges associated with technological adoption and digitalization.

This study aimed at preparing three types of biochar derived from rice husk via pyrolysis, including unactivated biochar, biochar chemically activated after with H₃PO₄, and biochar impregnated with γ-Fe₂O₃ nanoparticles. These materials were subsequently characterized using Scanning Electron Microscopy (SEM), Fourier Transform Infrared Spectroscopy (FTIR), and Brunauer–Emmett–Teller (BET) analysis, which revealed favorable textural properties, such as an increased surface area and porosity, as well as the presence of functional groups that facilitate the adsorption of methylene blue and malachite green in aqueous solutions. Several factors that affect the adsorption capacity, including the type of material, pH effect, and adsorbent dosage, were evaluated and optimized. The adsorption behavior was analyzed using isotherm and kinetic models to better understand the mechanisms involved. Under optimal conditions, biochar@γ-Fe₂O₃ NPs emerged as the most effective material due to its high surface area, functionalized surface, and magnetic properties, allowing easy water recovery without the need for complex instrumentation. Among the kinetic models evaluated, the pseudo-second-order model exhibited the highest linear regression coefficient (R² = 0.99), supporting a chemisorption process driven by strong interactions and stable chemical bond formation between the adsorbate and the adsorbent, while equilibrium data fit well with the Sips isotherm model, indicating a combination of monolayer and multilayer adsorption mechanisms. This magnetic biochar achieved removal efficiencies of 97% for methylene blue and 95% for malachite green, demonstrating a high performance and reusability over four cycles. Moreover, a possible adsorption mechanism of MB on the magnetic biochar was proposed to explain the interaction between the dye and the adsorbent surface. Thus, this work demonstrates that magnetic biochar is a sustainable and cost-effective adsorbent for wastewater treatment, integrating circular economy principles by transforming rice husk into a high-value material. The incorporation of γ-Fe₂O₃ nanoparticles enhances adsorption while enabling magnetic recovery, providing an eco-friendly and scalable solution for dye removal.

Este estudio analiza la valorización del pseudotallo del banano como una alternativa eco-sustentable en Ecuador, destacando su potencial para generar materiales biodegradables y fortalecer la economía circular. A través de una metodología cuantitativa basada en encuestas a 197 productores bananeros de la Zona 5, se exploraron percepciones y prácticas sobre el manejo de este subproducto. El análisis de red reveló que preguntas clave, como la disposición para comercializar el pseudotallo y el conocimiento sobre sus aplicaciones, actúan como nodos centrales en las percepciones de los productores. Los hallazgos sugieren una fuerte conexión entre la capacitación técnica y la adopción de prácticas sostenibles. Además, se identificaron barreras relacionadas con el acceso a tecnologías y mercados, lo que resalta la necesidad de políticas públicas que faciliten la implementación de estrategias de valorización. Este trabajo contribuye al desarrollo de modelos productivos sostenibles, promoviendo la integración de los pequeños productores en cadenas de valor eco-innovadoras.

This study evaluated the effect of thermal aging on the tribological properties of biocomposites formed by an isotactic polypropylene matrix (iPP) reinforced with 20 wt% (PP/20F), 30 wt% (PP/30F), and 40 wt% (PP/40F) of randomly oriented bamboo fibers. iPP, along with the grafting of maleic anhydride molecules (MAPP), was used as a coupling agent. The accelerated thermal aging involved the continuous heating of the materials at 98 °C for 10 days. Wear tests were performed under the Pin-on-Disk configuration to determine the wear factor (K) and the friction coefficient (µ) of the materials. After thermal aging, the µ value of the PP/20F composite increased by 40.5%, while, for raw PP, PP/30F, and PP/40F, the increase was 2.1, 7.5, and 2.2%, respectively. The aged PP/30F composite achieved the highest µ value. The loss of wear resistance due to aging was more prominent in the raw PP. The K factor of the aged and unaged PP/20F was the lowest. The use of scanning electron microscopy allowed identifying that adhesive, abrasive, and fatigue wear mechanisms were the dominant ones.

In this article, a characterization of the scientific production and editorial management of the journal Ciencia Unemi was carried out, where the universe of study is composed of its publications in Web of Science (WoS). The objective is to analyze the editorial management of the journal Ciencia Unemi in the period 2008 to 2022 using the bibliometric analysis of production, document typology, language capacity, authorship and collaboration, emphasizing the identification of the main research topics, their relationship and the components that integrate them. The study was descriptive in scope, longitudinal in trend (analyzing changes over time) and retrospective, using quantitative techniques and supported by the theoretical and methodological framework provided by the metric studies of information. To characterize the scientific production of the journal Ciencia Unemi, Web of Science (WoS) was used, since it offers indicators about the trends shown by countries, institutions, disciplines, journals and authors, in terms of production, collaboration and use of the work. It is concluded that the journal constitutes a reference of internationalization and visibility in Ecuadorian science, it is indexed in national and international databases, given the important presence of external production, whose performance drives the growing collaboration among researchers and the downloading of articles carried out predominantly outside national borders. However, this effort must be translated into citations to the journal, where there is a pending issue and several areas of opportunity and strengthening. En este artículo se realizó una caracterización de la producción científica y gestión editorial de la revista Ciencia Unemi, donde el universo de estudio lo componen sus publicaciones en Web of Science (WoS). El objetivo de esta investigación es analizar la gestión editorial de la Revista Ciencia Unemi en el periodo 2008 a 2022 usando el análisis bibliométrico de producción, tipología documental, capacidad idiomática, autoría y colaboración, haciendo énfasis en la identificación de los principales temas de investigación, su relación y los componentes que los integran. Se realizó un estudio de alcance descriptivo, longitudinal de tendencia (analiza cambios a través del tiempo) y retrospectivo, usando técnicas cuantitativas y sustentado en el marco teórico y metodológico provisto por los estudios métricos de la información. Para hacer la caracterización de la producción científica de la revista Ciencia Unemi se recurrió a Web of Science (WoS) dado que ofrece indicadores acerca de las tendencias que muestran los países, instituciones, disciplinas, revistas y autores, en términos de producción, colaboración y uso de la obra. Se concluye que la revista constituye un referente de internacionalización y visibilidad en la ciencia ecuatoriana, está indizada en bases de datos nacionales e internacionales, dada la importante presencia de producción externa, cuyo desempeño impulsa la creciente colaboración entre investigadores y la descarga de los artículos llevada a cabo predominantemente fuera de las fronteras nacionales. Sin embargo, este esfuerzo debe traducirse en citas a la revista, donde hay una asignatura pendiente y diversas áreas de áreas de oportunidad y fortalecimiento.

In the present investigation, the effects of fiber content (20, 30 and 40 wt%) and annealing heat treatment on the thermal properties of the biocomposite material formed by an isotactic polypropylene polymeric matrix (iPP) reinforced with short bamboo fibers (PP/BF) were studied. The coupling agent was iPP with grafting of maleic anhydride molecules (MAPP). The Differential Scanning Calorimetry (DSC) technique was used to determine the melting temperature (Tm), crystallization temperature (Tc), and degree of crystallinity (Xc) of the biocomposites. The melting temperature of pure PP was not significantly affected by the presence of bamboo fibers. However, the degree of crystallinity of PP increased with fiber content. The annealing heat treatment caused an increase in the Tm of pure PP and PP/BF. The composite with 20 wt% fiber content (PP/20F) annealed achieved the highest Xc (37,47 %). The Tc of polypropylene increased with the presence of fibers, and the PP/20F composite reached the highest value (115 degrees C). Also, the study of the Tm of the biocomposites in amorphous state was carried out, and it was observed that the sample with a fiber content of 30 wt% presented an endothermic peak at a temperature of approximately 168 degrees C.