Programa del congreso

The review analyzed how Artificial Intelligence optimized candidate identification, assessment, and selection processes using a methodology based on 21 articles selected through the PIOC strategy, specific inclusion and exclusion criteria, and the PRISMA process, ensuring rigor in identifying relevant evidence. The results showed that AI improved operational efficiency, assessment accuracy, and the detection of performance patterns not visible through traditional methods, although challenges related to algorithmic transparency, bias, perceived fairness, and technological acceptance persisted among both candidates and recruiters. The discussion and conclusions indicated that the effectiveness of these tools depended on their ethical, technical, and organizational integration, highlighting that AI should function as a complement to human judgment and projecting the need for future research to strengthen its responsible use in selection processes.

Disabilities affect 1.3 billion people (16% of the world's population) [1], with upper limb motor limitations being a leading cause of functional dependence and generating high demand for specialized rehabilitation. This article presents the development of NERO, a rehabilitation and assistance platform based on a multimodal brain-computer interface (BCI) that integrates EEG signals, EMG signals, and inertial direction validation using MPU sensors. The system enables intuitive control of an arm-support exoskeleton through directed mental and muscular patterns and sustained head movements. A threshold-based activation attention logic combined with timing is implemented, ensuring safety and personalization. The architecture is complemented by a real-time visualization interface using Apache Superset, connected to a PostgreSQL database hosted in Docker and backed up by a local database. Nero offers an accessible, modular, and replicable solution in the field of therapeutic neurorobotics, democratizing access to low-cost neuromotor rehabilitation technologies and prosthetic control.

According to the Latin American and Caribbean Energy Organization (OLACDE) (2023), Latin America and the Caribbean generated more than sixty percent of their energy from renewable sources. However, this progress introduces new challenges related to the end-of-life management of technologies such as solar panels and wind turbines. In this context, extending operational lifespan is a priority, and digital twins emerge as a key tool for promoting industrial sustainability; these virtual representations enable the simulation, monitoring, and optimization of energy systems in real time, extending their service life and reducing operational costs by facilitating predictive maintenance. Furthermore, they contribute to the achievement of the Sustainable Development Goals, particularly SDGs 7 and 12.

This research presents a bibliometric and systematic review on the application of digital twins in renewable energy, structured around specific research questions. To this end, databases such as Scopus, Lens, and Elsevier were consulted using defined selection criteria.

The results will highlight the potential of digital twins to improve energy sustainability in the region, offering innovative solutions tailored to the Latin American context.

In the current landscape of educational digital transformation, integrating new technological tools to enhance student learning poses significant challenges. Notably, following the COVID-19 pandemic, potential cognitive and attentional delays have been observed in children due to disruptions in traditional learning processes. Therefore, developing mechanisms to accurately measure the impact of these educational tools is crucial since existing evaluations mainly rely on subjective behavioral observations. A notable gap exists in the literature regarding their effectiveness and neurological validation. As the validation phase of a broader research project, this study aims to develop a tool for objectively assessing cognitive states. The methodology involves conducting validated tests with children, recording Electroencephalogram (EEG) signals, and applying predictive machine learning models such as Random Forest to classify cognitive states during interactions with the NAO robot, a Socially Assistive Robotics (SAR) platform. Our preliminary results demonstrate the feasibility of classifying students' attention states—engaged versus distracted—with high accuracy (79%). This approach provides a robust quantitative metric that complements behavioral observations, offering educators and researchers a data-driven tool to evaluate and personalize teaching strategies, thereby helping to close learning gaps. Ultimately, this study advances the integration of educational neuroscience, artificial intelligence, and SAR in educational settings.