Conference Agenda

Gender stereotypes continue to shape educational choices and career trajectories, contributing to persistent underrepresentation of women and other marginalized groups in engineering. LaleoLab, a research-informed board game originally developed in Italy, offers an innovative approach to addressing these challenges through collaborative, game-based learning. By engaging students in critical reflection on gender roles and their influence on STEM pathways, LaleoLab fosters inclusive thinking and challenges normative assumptions about who “belongs” in engineering.

This paper reports on the adaptation and implementation of LaleoLab within UK higher education, highlighting its potential as a pedagogical intervention for promoting diversity in engineering. Building on a multicultural student workshop funded by a Society & Culture Seed Grant at the University of Warwick, the game was enriched with additional “special cards” featuring diverse role models—including scientists, engineers, and innovators from the Global South—responding to student feedback that representation must extend beyond Eurocentric narratives. These additions aim to broaden students’ understanding of engineering as a global discipline and to inspire aspirations among underrepresented groups.

We discuss the educational impact of LaleoLab in fostering critical awareness, intercultural dialogue, and engagement with equity issues in engineering education. Preliminary findings suggest that integrating playful, reflective activities into curricula can complement traditional teaching, supporting efforts to dismantle stereotypes and create more inclusive learning environments. Future work will explore classroom implementation and evaluate its influence on students’ attitudes toward engineering careers.

Mining operations often generate a range of waste and effluents that must be managed and disposed of safely for company personnel, surrounding communities and the environment. In this sense, a mining company must implement an environmental management system in accordance with Peruvian legislation. This paper shows the methodology to follow to implement an environmental management system that can resolve and prevent potential environmental impacts already generated or about to be generated. For the case study, an underground mining operation was chosen that is one of the world's leading producers of zinc, silver and lead. This methodology includes the following steps: 1) developing the company's environmental commitment and policy, 2) planning the activities to be developed at each stage of the production process, 3) implementing the management system, 4) measuring and evaluating the implemented system, 5) proposing a continuous improvement plan.

A major source of water pollution from mining is acid mine drainage, produced by the oxidation of minerals with high sulfide content. Currently, the possibility of using bioadsorbents to remove heavy metals has become an attractive alternative. This article presents preliminary results on the removal of copper and cadmium using grape waste from the wine industry. Through batch removal trials, optimal removal conditions will be established by monitoring the amount of biosorbent, effluent volume, contact time, and temperature. The final concentration of each contaminant will be determined using atomic absorption spectroscopy (AAS). The maximum removal percentage will also be determined for each contaminant. Removal tests demonstrate that it is possible to remove up to 71% of copper and 34% of cadmium. The Langmuir model fit better than the Freundlich model for copper (R² = 0.876) and cadmium (R² = 0.957) adsorption in grape pomace. The aim is to use waste from the wine industry as an efficient, economical, and eco-friendly bio-additive, making it an attractive and competitive alternative to other contaminant removal technologies for mining effluents.

This study examines the correlation between the implementation of **Problem-Based Learning (PBL)** as a didactic strategy in disciplinary courses of the Civil Engineering program at Areandina and its impact on the development of specific competencies, as evidenced by the results of the *Saber Pro* examinations administered by ICFES, the agency designated by the Colombian Ministry of National Education to assess educational quality. The research arises from the low academic performance and below-average scores obtained by students in specific competencies, highlighting the need to strengthen the teaching–learning–assessment triad. A quantitative approach with a correlational design was adopted, grounded in the systematic application of PBL, and conducted with a sample of 315 students and nine faculty members between 2020 and 2024, selected through non-probabilistic convenience sampling. The findings reveal a positive correlation between the use of PBL and the continuous improvement of both student scores in the *Saber Pro* examinations and learning curves, attributable to a 12% increase in the strengthening of specific competencies. The study concludes that the transversal incorporation of PBL into disciplinary courses is indispensable for optimizing professional training in Civil Engineering.

This paper examines cognitive friction in digital cognitive assessment workflows for older adults, focusing on both the patient experience and the clinical staff administering standardized tests. Grounded in Cognitive Load Theory and classical usability principles, we position AI as cognitive scaffolding that reduces extraneous load in two layers: software development and bedside execution. We propose a practical low-friction design guide for tablet-based assessments, emphasizing explicit affordances, single-path navigation, multisensory feedback, and objective trace capture. Using early field experience from Parkinson’s disease screening brigades in Colombia, we illustrate how automated scoring, voice capture, and reaction-time logging improve operational reliability and data quality without increasing perceived complexity. We conclude that “invisible interfaces” are not only a performance choice but an ethical requirement to expand access and reduce technology-related anxiety in vulnerable populations.

This paper presents the progress in the development of an autonomous robotic system for strawberry harvesting, based on computer vision and deep learning techniques. It addresses the mechanical design of a robotic platform with three degrees of freedom, as well as its conceptual integration with a perception system composed of an RGB camera, a depth sensor, and an embedded system. Additionally, the modeling of the prototype and its incorporation into a simulation environment in CoppeliaSim are presented as a preliminary step to the physical validation of the system. For the perception stage, a classic computer vision approach was implemented, combining image segmentation techniques with convolutional neural network models for classifying fruit into harvestable and unharvestable states. Different neural network architectures were evaluated, achieving performances exceeding 95% accuracy in a controlled environment, demonstrating the system's potential for fruit sorting applications. Finally, the foundations for future stages of the project are established, which include the integration of the perception system with the robot's movement, the validation of the system in more realistic environments, and the comparison with end-to-end approaches based on detection through deep learning.

This paper investigates the culture of energy efficiency in household energy consumption, focusing on the community of the Universidad Nacional de La Matanza (UNLaM). Within the framework of the energy transition, the study argues that technological and regulatory solutions are insufficient if sociocultural dimensions and users’ cognitive capital are not adequately integrated. The central hypothesis suggests the existence of a deficit in energy literacy and an underestimation of cultural factors within current dissemination and communication programs. Methodologically, the research adopts a descriptive and exploratory approach based on a structured survey, assessing eight analytical dimensions, including social norms, infrastructure, and implementation barriers. The main objective is to analyze the gap between theoretical knowledge and everyday energy consumption practices. The expected results aim not only to validate the proposed hypothesis, but also to contribute empirical evidence to support the design of more effective and context-sensitive public policies, regulations, and communication programs.

This study examines the integration of serious game development within an undergraduate systems and software engineering curriculum using a Project-Based Learning (PBL) approach. Engineering students engage in the design and implementation of a narrative-driven serious game addressing the armed conflict in Colombia as a socially grounded application context. The development activity is framed as a system-oriented engineering task, involving architectural design, modular software development, interaction modeling, system integration, and iterative refinement under realistic technical and organizational constraints. The work focuses on engineering learning processes associated with sustained engagement in the development of a complex software system. It is organized around multiple development phases, including system analysis, architectural definition, component implementation, and integration activities. These phases provide a structured context in which students engage with system decomposition, interface definition, state management, and technical trade-offs across the software development lifecycle. By embedding the development task within the formal curriculum, the project supports extended engagement with open-ended engineering problems that require coordination among interdependent system components and adaptation to evolving requirements. The socially grounded narrative context introduces additional design considerations related to representation, interaction sequencing, and system behavior, which must be addressed through architectural and software engineering decisions. This paper contributes to understanding how complex, system-level software development projects situated within undergraduate engineering education can support systems thinking and the development of core engineering competencies.

Intelligent resource management in agriculture reduces costs and maximizes production yields with minimal inputs, contributing positively to environmental sustainability. New technologies based on digital agriculture allow for estimating the quantity, location, and timing of agricultural input applications according to the actual needs of plants and soil. Current technologies such as the World Wide Web, the Internet of Things (IoT), sensor networks, and artificial intelligence are widely used in agriculture. The purpose of this work is to demonstrate the progress in developing a web platform based on Industry 5.0 technologies that supports irrigation management in agricultural crops and their adaptation to climate change by leveraging the existing relationships between the environment, plants, and soil. The following sections detail the system architecture design, the development of data acquisition modules, remote information management, and its implementation in a real-world environment. The work developed is a tool that supports intelligent decisions, helping farmers to take actions in the field based on information.

Artificial intelligence (AI) and virtual collaboration are reshaping professional practice, yet many postgraduate curricula still treat AI literacy, internationalisation, and career development as separate strands rather than integrated initiatives. This work-in-progress reports on a quasi‑experimental, mixed‑methods study connecting Monash University Malaysia (Advanced Career Counselling unit) and the University of Warwick (Fundamentals of AI Research, Development and Management module). Approximately 210 master’s students are allocated into four conditions: (A) control; (B) virtual, interdisciplinary exchange; (C) AI‑supported presentation training (Microsoft Speaker/Presenter Coach); and (D) combined exchange & AI training.

The study evaluates impacts on engagement, skill development, intercultural competence, digital fluency, and career preparedness. The study uses pre/post-surveys, rubric‑based assessments, and interviews/focus groups. Ethical safeguards address power imbalances (students of the investigators), explicit consent, data minimisation, and cross‑border data handling. Early implementation experiences, instrumentation, and planned analyses are presented to inform future interdisciplinary international curricular collaborations and AI-supported assessment.

Digital agriculture has significantly enhanced productive efficiency through the intensive use of technologies such as the Internet of Things (IoT), Big Data, artificial intelligence, and cyber-physical systems; however, its increasing energy demand poses substantial environmental challenges. This paper examines the role of Green Computing as a technological cornerstone for the development of sustainable digital agriculture, aimed at reducing energy consumption and carbon footprint without compromising productivity. The adopted methodology consists of an analytical and systematic review of recent scientific literature, complemented by the examination of documented case studies and a comparative analysis of green technologies applied to the agricultural sector. The findings indicate that the integration of edge and fog computing architectures, Green Machine Learning algorithms, low-power IoT sensors, and renewable energy sources enables significant reductions in water use, agricultural inputs, and fossil energy consumption, achieving decreases of up to 80% in conventional energy use and productivity improvements ranging from 10% to 30%. The study concludes that the adoption of green computing principles represents a critical strategy for ensuring environmental sustainability, operational resilience, and the long-term viability of digital agriculture, while aligning with the Sustainable Development Goals and addressing the challenges posed by climate change.

The integration of artificial intelligence (AI) into higher education presents a strategic opportunity to transform teaching and learning processes in engineering, particularly in Latin American contexts facing structural challenges related to equity, student retention, and pedagogical innovation. This paper presents a mixed-methods study conducted at the National Technological University, Delta Regional Faculty (UTN–FRD), aimed at analyzing the impact of AI tools on engineering student learning, explicitly integrating a gender perspective and an ethical approach.

The study examines the implementation of intelligent tutors, content recommendation systems, learning analytics, and adaptive assessments in selected courses from the basic and advanced cycles. Variables related to academic performance, motivation, self-regulated learning, and perceived equity are analyzed. The expected results aim to provide empirical evidence to support the responsible integration of AI as a driver of educational innovation and as a tool for reducing gender gaps.

Small-scale rice farmers in resource-constrained rural regions experience persistent productivity losses due to limited access to advanced agricultural technologies for pest and disease management. This paper presents a work-in-progress study on the development of an offline software system integrating Augmented Reality (AR) and Artificial Intelligence (AI) to support real-time detection of pests and diseases directly in the field. The proposed system combines computer-vision-based detection, an AI-driven chatbot, and AR visualization through low-cost AR glasses, enabling farmers to receive actionable recommendations without continuous internet connectivity. The project follows a phased methodology encompassing data collection, prototype development, field testing, and system refinement. At the current stage, a balanced dataset of 2,000 rice images captured using Vuzix Blade 2 AR glasses has been annotated and used to train and test a YOLO11n-based object detection model in a desktop environment, producing preliminary qualitative detection and localization outputs. This work aims to improve crop management efficiency, reduce chemical inputs, and enhance agricultural sustainability, while establishing a foundation for future quantitative evaluation and large-scale deployment.

Rural education systems and smallholder farming communities in Colombia and across Latin America face parallel challenges related to agricultural sustainability, technological renewal, and long-term generational continuity. In rural schools, limited opportunities for applied technological engagement weaken students’ capacity to connect agricultural education with future career pathways, while among small and medium-scale producers, the adoption of Agriculture 4.0 technologies remains limited. Although tools such as Internet of Things, drone-based sensing, and augmented reality offer benefits for productivity and environmental management, their lack of a significant presence in rural contexts underscores the need to better understand how such technologies are perceived, interpreted, and evaluated within everyday educational and farming practices. In response, this paper defines an adoption-oriented analytical framework for examining how Agriculture 4.0 technologies are introduced and assessed across rural agricultural education and smallholder farming settings. The framework is structured around a two-phase process that first examines perception formation following contextualized technological exposure and subsequently analyses evaluative judgment emerging through applied interaction in settings that approximate real agricultural decision-making. The department of Meta, Colombia, is designated as the intended application context for the framework, although full implementation is beyond the scope of the present study. Preliminary results draw on baseline surveys, interviews, and early qualitative observations to characterize initial adoption-related perceptions and to validate the readiness of the technological and pedagogical components supporting future framework execution. The proposed framework establishes a structured methodological basis for continued study and is intended to be transferable to other rural regions with comparable educational and socio-productive conditions.

Water resource management in agriculture is crucial for preventing yield losses and ensuring the most efficient use of resources. One irrigation prescription mechanism is the early and precise determination of plant water stress in a crop, enabling targeted irrigation. Infrared sensors measure the temperature of the plant canopy, reflecting the relationship between radiation, transpiration, and plant water status. This document presents the progress in the design and implementation of a prototype data acquisition system using electronic devices to determine plant water stress in an agricultural crop, based on leaf temperature measurements. The system acquires data and transmits it to a central station for monitoring and, in the next stage of the project, to define irrigation prescriptions using a deep learning model (LSTM). These advancements demonstrate the applicability of the developed system to irrigation management in agricultural crops.

The educational process undertaken by students creates an ideal scenario for studying the educational trajectories of programs through longitudinal analyses. This type of analysis allows for the evaluation of trends and patterns in students' academic progression, as well as low graduation rates, the time taken to complete their professional studies, and high dropout rates. Each of these aspects should be studied by institutions to improve program quality, which is assessed through accreditation and registration processes.

In Colombia, the average dropout rate at the higher education level is approximately 45%, while in Latin America, it is 50%. For this study, a mixed-methods approach is proposed. The quantitative approach is supported by the application of descriptive and correlational techniques, as well as longitudinal models, based on the analysis of information provided by university programs.

Bats use ultrasonic echolocation signals to orient themselves and feed, and analyzing these signals can be used for non-invasive, acoustic monitoring. Traditionally, this process is carried out manually by experts, which is a slow process due to the large volume of data and the temporal resolution of the signals. Currently, acoustic monitoring is an essential tool in ecology, especially for the study of bats, due to its non-invasive nature and its broad spatial and temporal coverage. For this reason, artificial intelligence techniques, including deep learning, have been incorporated into this work. This paper presents the use of region-based convolutional neural networks using a Faster R-CNN model for the automatic detection of pulse sequences from bat species in ultrasound recordings from spectrograms. A preliminary model showed high completeness values (>89%) and solid performance at mAP@50 (>892%), demonstrating adequate spatial localization capacity for the sequences, even in high biodiversity environments. However, lower accuracy and confusion values were observed between acoustically similar species, probably due to simultaneous vocalizations and the limited diversity of the training set. The feasibility of this approach for automated acoustic monitoring is evident, and future work should focus on expanding the dataset and improving interspecies discrimination.

The COVID-19 pandemic has been one of the most severe public health crises of recent decades, exerting profound impacts on population health, economic activity, and social dynamics.
In Chile, systematic epidemiological surveillance conducted by the Ministry of Health (MINSAL) during 2020--2021 yielded an anonymized dataset of 54 variables and 6,764,619 patient records.
This study aims to infer an interpretable causal structure that relates demographic, clinical, and health-system-related variables to COVID-19 outcomes using observational surveillance data.
We learn a directed acyclic graph (DAG) with Greedy Equivalence Search (GES) and evaluate its plausibility using a permutation-based falsification test in DoWhy, which quantifies violations of the Local Markov Condition (LMC) relative to randomized baselines.
The learned GES graph is informative under falsification (0/20 permutations in the Markov equivalence class; p-value = 0.00) and exhibits fewer LMC violations than permuted graphs (79/1629), indicating that it encodes characteristic conditional independencies unlikely to arise by chance.
Overall, the resulting structure provides a compact and interpretable causal hypothesis space to support domain interpretation and to prioritize variables and pathways for subsequent causal effect analyses.

Wildfires in central Chile have intensified in frequency and impact, yet operational models often remain limited to correlational predictors, hindering robust interpretation and decision support. This study aims to discover and validate the causal structure underlying wildfire occurrence in the Maule Region, Chile, identifying the most plausible drivers and their directed dependencies. We compile a multi-source dataset integrating wildfire records with meteorological, vegetation, topographic, and anthropogenic variables, harmonized to a common spatiotemporal grid. Causal discovery is performed using constraint-based and score-based structure learning under a linear non-Gaussian assumption, complemented with stability selection and sensitivity analyses to assess robustness against sampling variability and confounding. The inferred graphs consistently recover a sparse, stable causal backbone in which short-term meteorological conditions and fuel-related proxies act as primary upstream drivers, while human-accessibility variables exhibit mediated effects through ignition likelihood. Across multiple resampling runs, key directed edges remain stable, and the resulting causal model improves out-of-sample interpretability and risk attribution relative to purely associative baselines. This causal framework provides actionable insight into drivers likely governing wildfire dynamics in Maule, enabling more defensible early-warning indicators, targeted prevention policies, and a principled basis for scenario analysis through interventions on controllable factors.

Wildfires have become an increasingly recurrent hazard in central Chile, producing severe environmental and socioeconomic impacts and motivating probabilistic, spatially explicit tools for prevention and operational planning. Focusing on the Maule Region, this study estimates wildfire occurrence probability in high-risk zones using a spatiotemporal modeling approach that captures both local environmental conditions and their temporal evolution. We compile and harmonize a gridded dataset combining wildfire occurrence records with dynamic drivers (e.g., meteorological variables and vegetation-related proxies) and static descriptors (e.g., topography), and we formulate risk estimation as a sequence-to-probability task at the grid-cell level. We develop a Transformer-based spatiotemporal architecture that learns nonlinear dependencies across time and space and outputs calibrated probabilities of wildfire occurrence for multiple forecast windows. Performance is assessed on held-out periods using discrimination and calibration analyses, and the predicted probabilities are converted into risk-zonation maps to support consistent operational thresholding. Results show that the proposed model produces coherent probability fields that concentrate risk in historically fire-prone areas under conducive conditions, while delivering stable probabilistic outputs suitable for decision-making. The resulting risk maps provide actionable guidance for prioritizing surveillance and allocating prevention and response resources in the Maule Region, and the framework is readily transferable to other regions where wildfire risk exhibits strong spatiotemporal variability.

The present project was developed to propose alternatives for the utilization of waste generated during the wood transformation process in a commercial Pinus patula forest plantation located in Boyacá, Colombia. These residues, traditionally stored without a defined use, present significant potential for valorization. To this end, engineering tools and quantification methods were applied, making it possible to determine that, out of more than 16,000 tons of annual waste, 60% of the offcuts and 85% of the sawdust and wood shavings can be reused. Among the proposed strategies, the production of wooden slats and the direct commercialization of by-products stand out, both of which demonstrate economic viability with estimated annual revenues of COP $248,500,000, in addition to contributing to the reduction of environmental impact and the strengthening of sustainable practices within the Colombian forestry sector.

Solid waste management is an ongoing challenge in the city of Chiclayo, particularly due to the limited integration of informal waste collectors into the municipal system. The objective of this research was to analyze the main challenges facing the formalization of informal solid waste collectors in Chiclayo during the year 2026. The study was conducted using a basic qualitative approach, with an exploratory-descriptive design and a phenomenological case study. Information was collected through semi-structured interviews with formal and informal waste collectors, municipal authorities, and environmental management specialists. The results show that the main barriers to formalization are bureaucratic complexity, limited regulatory knowledge, the perception of high legal costs, and mistrust of public institutions. It is concluded that formalization requires simplified processes, effective incentives, and greater institutional coordination to ensure decent working conditions and sustainable solid waste management

The Terra-Digital Twin project represents a paradigm shift in environmental monitoring. The proposal integrates ephemeral hardware sensors (biodegradable) with a mesh network infrastructure and AI-based orchestration. Unlike current solutions, this system creates a dynamic virtual replica (Digital Twin) of terrestrial ecosystems without generating electronic waste. The goal is to provide governmental and private managers with a high-resolution climatic and biological prediction tool, optimizing decision-making through deep learning models and sustainable technology management.

The proposal consists of providing scientific and technological mentoring to high school students in order to generate interest and involvement in scientific research activities and projects, and thus motivate them to continue their studies in STEAM higher education programs.

This project involves a constructivist intervention based on the 5E methodology, led by two key figures: research staff (full-time research professors) and university mentors (higher education students). The research laboratories of the university itself will be used in the project.

The purpose of this research is to find out if female students studying upper secondary education were motivated by the intervention of scientific and technological mentoring to continue their studies in STEAM engineering programs.

In veterinary medicine, the study of histology and pathology is essential for understanding normal and pathological biological processes, as well as for developing key diagnostic skills in oncology and pathological anatomy. However, access to physical histological preparations depends on specialized infrastructure, the condition of the samples, and teaching support, which limits independent learning and open access to knowledge. The digitization of histological slides using Whole Slide Imaging (WSI) technology allows some of these barriers to be overcome, replicating the microscope experience with high fidelity. However, the size of the files poses challenges for storage, organization, and visualization, especially in institutions with limited technological resources.

To address this problem, a data model and architecture were developed, and mockups of a digital atlas for veterinary pathology practice were created. This tool will enable the efficient uploading, storage, and visualization of high-resolution histological samples, integrating navigation, search, and interactive tagging functionalities by systemic classification.

ZSM-5 zeolite was synthesized through a seed-assisted hydrothermal method as a sustainable alternative to conventional OSDA-based routes and evaluated for ethanol dehydration to ethylene. The synthesis employed tetraethyl orthosilicate as silicon source, sodium aluminate as aluminum source, and commercial ZSM-5 seeds with different Si/Al ratios. The influence of crystallization time on morphology and phase purity was examined by SEM and X-ray diffraction. Well-defined MFI crystals were obtained after 72 h of hydrothermal treatment, while longer synthesis times led to the appearance of minor extra-framework silica phases. The catalytic performance of the H-ZSM-5 samples was evaluated in a fixed-bed reactor between 150 and 550 °C. The catalyst exhibited clear temperature-dependent behavior, achieving moderate ethanol conversion and ethylene formation at 550–500 °C, confirming the presence of active Brønsted acid sites. However, incomplete conversion at the highest temperature suggests limitations associated with acid site density and/or mass-transfer constraints. Overall, the seed-assisted route produced active ZSM-5 catalysts while avoiding organic templates, demonstrating potential as a more sustainable pathway for ethylene production from bioethanol.

This paper analyzes resource management and strategies for the export performance of small and medium-sized enterprises (SMEs). Using a mixed method approach that integrates bibliometric analysis of 166 documents with Bibliometrix of RStudio and a multiple-case study involving in-depth interviews with five exporting SMEs, analyzed through NVivo. Findings reveal patterns in scientific productivity and empirical evidence on how competitive advantages, business environment, and marketing strategies can influence profitability, internationalization, and firm performance.

Optimizing crops under cover in high Andean regions requires precise control of climatic factors. This article presents the design and implementation of an automated system for the monitoring and control of temperature, relative humidity and soil humidity in greenhouses located in Pamplona, ​​Norte de Santander. The system integrates a central node based on the ESP32 microcontroller, precision sensors and a wireless communication architecture. The results show that automation stabilizes the internal microclimate and optimizes the water resource by 28% through data-based management.

Automated analysis of scientific literature represents an opportunity to discover relevant knowledge in a specific domain. The main challenge of Natural Language Processing (NLP) is the morphological descriptions of technical language and their scattered location, which hinders systematic processing.

This work proposes the design and implementation of an NLP method such as bidirectional encoder representation from transformers (BERT) for the automatic analysis and visualization of information extracted from scientific articles on artificial intelligence technologies used to support veterinary pathology. Techniques such as Named Entity Recognition (NER) and semantic representations are used, supported by pre-trained models from the scientific domain. The work was approached from the construction of a specialized corpus, text processing, fine-tuning of the BERT and SciBERT models, and evaluation using standard metrics. In the first experiment, the best configuration obtained in fine-tuning was SciBERT_lrst1, which achieved an F1-score of 0.8262, with a learning rate of 5e-05, weight decay of 0.1, 100 warmup steps, and a linear scheduler. The second experiment maintained the same hyperparameters and the best model, BERT_lrst1, obtained an F1-score of 0.7573. It can be observed that the incorporation of warmup steps and a higher level of regularization promote training stability and improve generalization capacity, which is particularly relevant in scenarios with limited corpus size. SciBERT's specialization in scientific texts gives it an additional advantage in NER tasks in specialized domains such as the one analyzed in this study.

This systematic scoping review establishes the evolving trends of AI within engineering education. Analysis of 2,500 publications identifies a decisive shift from "detection" to "integration". We demonstrate a persistent Theory-Practice Gap, where technical implementation outpaces pedagogical anchoring The rapid emergence of Large Language Models (LLMs) has created a "literature deluge," where the speed of technological adoption often outpaces the synthesis of evidence-based practices. This study analyzes approximately 2,500 publications from the ASEE and FIE conferences to identify where AI application is moving and which pedagogical strategies are gaining traction. Key findings indicate a significant shift from 2024’s focus on "detection and restriction" to 2025’s focus on "integration and prompt engineering”.

The increasing production of cocoa in Colombia is generating a large amount of agro-industrial residues, especially cocoa pod husk (CPH) and cocoa bean shell (CBS). Without proper waste management, these residues may cause environmental problems such as soil contamination and greenhouse gas emissions. This study focuses on the energy valorization of CPH through pyrolysis, which is presented as an eco-friendly alternative for bio-oil production. The biomass was characterized prior to pyrolysis through proximate and elemental analyses. It was then subjected to thermal pyrolysis in the range of 550–700 °C, heating rates of 10 and 100 °C/min, 140 ± 5 mL/min of N₂ (carrier gas), and isothermal reaction times of up to 10 min. Under these conditions, bio-oil yields were calculated. Subsequently, the bio-oil will be evaluated using gas chromatography coupled with mass spectrometry (GC–MS) to estimate its calorific value through empirical models reported in the literature. The results will provide new insights into residues from the Colombian cocoa sector, offering valuable information to strengthen energy recovery pathways, promote the circular economy, and support future studies in agro-industrial sustainability

This study analyzes the evolution of the energy matrix and carbon intensity in Latin America and the Caribbean (LAC) during the 2000-2023 period, including a specific sub-analysis of the 2020-2023 interval. The research focuses on nine structurally significant economies: Brazil, Mexico, Argentina, Colombia, Chile, Peru, Ecuador, Bolivia, and Uruguay. By integrating data from the International Energy Agency (IEA) and the World Bank, an analytical framework based on the Kaya Identity and the Logarithmic Mean Divisia Index (LMDI) decomposition method was applied to quantify the drivers of GHG emissions.
The results reveal a marked regional heterogeneity: while countries such as Uruguay and Brazil exhibit strong decoupling with carbon intensities below 40 CO2/TJ, economies like Mexico and Argentina maintain a rigid fossil fuel dependency with values exceeding 60 CO2/TJ. The LMDI analysis demonstrates that economic growth (Activity Effect) remains the primary driver of emissions, often outpacing progress in energy efficiency and energy mix decarbonization.
It is concluded that the region possesses consolidated leadership in renewable electricity but faces an incomplete transformation of its primary energy systems. The study identifies an urgent need to integrate new energy vectors, such as green hydrogen and electromobility, to overcome fossil inertia in hard-to-abate sectors and achieve absolute decoupling in the región

The Cigarette Company (CI), a leading manufacturer with over 50 years of experience, specializes in machine-made cigars and is recognized for its commitment to quality and sustainability. This study addresses inefficiencies in waste management and production processes within the BasicA cigar production line, which accounts for 85% of the company’s output. Key issues include unstandardized waste documentation, inconsistent data collection, and operational inefficiencies. By implementing standardized operating procedures (SOPs), digital waste logging systems, and enhanced material handling practices, the project achieved a 15% reduction in tobacco waste and a 5% improvement in operator utilization.

The Student workload is a growing concern in higher education, particularly in engineering, where increasing curricular demands may affect students’ self-regulation and time management. While Learning Analytics has proposed indicators derived from Learning Management System (LMS) data to characterize workload, these measures may not fully reflect how educational stakeholders understand and use such information. This work-in-progress study aims to identify student workload indicators in higher education and examine how they may support student self-regulation and instructional design. To this end, a workshop was conducted in Chile with 40 higher education managers and academics, organized into eight groups, using open and closed card sorting activities. In the open card sorting, participants generated and grouped indicators they considered relevant for supporting students. In the closed card sorting, LMS-based indicators were classified according to the moment of the academic period in which they were perceived as most useful. Preliminary findings show that time-related indicators were the most salient across groups, followed by personal, goal-oriented, activity-based, and perception indicators. Participants mainly associated system-based indicators with monitoring during the academic period, while comparative and perception-based indicators were also seen as useful for earlier decision-making stages. These results suggest that meaningful workload indicators should combine behavioral, contextual, and subjective dimensions.

Mentoring has been used as a strategy to close gender gaps in STEM. In this study, we present an analysis of the implementation of a mentoring program designed to encourage female secondary school students to choose STEM elective courses. The analysis focuses on the experiences of the mentors, who were university students who delivered four online mentoring sessions to female secondary school students. Our findings are organized around three axes: mentors’ perceptions, the feedback they provided on the program’s implementation, and the benefits and drawbacks mentors perceived from their own experience. Overall, mentors found the online format difficult to achieve full participation from mentees. At the same time, mentors considered that the program was implemented positively, given the support structure and the materials provided for each session. Finally, mentors reported substantial benefits from the experience, especially personal gratification, mutual personal enrichment, increased self-confidence, and the development of soft skills. In conclusion, the experience was meaningful for mentors, who were able to reinterpret their own school trajectories and valued the opportunity to be the kind of support they would have wished for at that stage.

The choice of a university career is a decision based on the information available at the time of making the decision and depends on multiple variables related both to the individual and to the surrounding context. According to the literature, factors influencing career choice include job expectations, the possession of required skills and economic resources, gender, personal preferences and interests, as well as family and sociocultural environment. Among the most demanded careers within STEM disciplines, it has been observed that in the Faculty of Engineering and Sciences at Universidad de La Frontera there are two programs with higher preference among students on the autism spectrum. This work explores the role of visual representation of professions through images as a potential factor influencing career choice, and analyzes its relationship with the decisions of autistic students, considering gender differences.

Curriculum analytics are emerging as a key tool to support higher education institutions in evaluating the attainment of learning outcomes and competencies within accreditation contexts. However, it is essential for their adoption to have a meaningful impact on teaching and learning processes to involve different educational stakeholders as key informants during the design and implementation stages. In this study, the usability and perceived usefulness of a curriculum analytics solution aimed at measuring and visualizing competency attainment for continuous improvement and quality assurance were evaluated. To this end, the cognitive walkthrough technique was employed to examine the perspectives of academic managers and faculty members from a prestigious Latin American school of engineering. The findings reveal challenges related to usability, faculty training, and data literacy as necessary conditions for effective implementation. Finally, future projections and lines of action derived from the lessons learned are discussed, with the aim of contributing to the strengthening of engineering programs that are advancing toward a culture of continuous improvement.

This paper presents the case study of a flagship program at the School of Engineering of the Pontificia Universidad Católica de Chile aimed at the internationalization and development of students interested in entrepreneurship. This initiative, called The Bridge, emerged within the context of a public policy framework and seeks to provide talented students with entrepreneurial potential an early transformative experience to foster cultural change—both in themselves and in their environments—while also shaping their professional trajectories. The paper describes the different elements of the program and evaluates its outcomes. The results show that program participants maintain and strengthen career paths related to innovation and entrepreneurship and, upon their return, make concrete contributions to strengthening both the internal and national innovation ecosystems.

In this Work-in-progress (WIP) paper, we present the background, conceptual framework, and preliminary results of an ongoing faculty development at a Chilean regional university. The program focuses on developing innovation and research competencies among the College of Engineering faculty by forming and supporting a community of practice that brings together members with diverse expertise and experience levels.

While curricular flexibility in engineering programs enhances student autonomy, it introduces significant complexity in academic planning for students and institutions. Ineffective decision-making regarding course sequences can hinder academic progression and increase the risk of dropout. Consequently, analyzing curricular trajectories is essential for supporting student success. This study explores the application of deep learning models to model course enrollment sequences at the School of Engineering of the [ANONIMIZED INSTITUTION], using a dataset of 11,729 students between 2013 and 2025. We compare two neural network architectures—Bidirectional Long Short-Term Memory (bi-LSTM) and Bidirectional Encoder Representations from Transformers (BERT)—against a K-Nearest Neighbors (KNN) baseline in a masked course prediction task. Our results demonstrate that BERT significantly outperforms the alternative models. This superiority is attributed to BERT's ability to model concurrent course enrollments within a single academic period through positional encoding. These findings suggest that BERT-based architectures are highly effective for understanding complex curricular trajectories, offering a robust foundation for course recommendation systems and student progression analytics. This approach opens new avenues for identifying "bottlenecks," predicting timely graduation, and mitigating dropout risk in flexible educational environments.