ABSTRACT. .

ABSTRACT. This paper presents a novel image processing algorithm for enhancing the quality of X-ray images, with the aim of improving the accuracy and efficiency of medical diagnosis. The algorithm combines two types of filters: the Cream filter, based on the work of Lefranc-Bustamante, and the Bosso filter, which uses a Gaussian function. These filters model the degradation in the quality of X-ray images due to the effects of radiation scattering and attenuation. The proposed algorithm was evaluated on three types of images: knee, breast, and wrist, and the results show a significant improvement in contrast and sharpness, particularly in knee and wrist images. The behavior of entropy and contrast as a function of the filtering parameter d was also analyzed, and the implications of these findings are discussed. The paper concludes by suggesting further explorations of the algorithm in different X-ray imaging techniques and the potential impact of these improvements on the accuracy of medical diagnosis.

ABSTRACT. This paper focuses in concrete at processes done at synaptic cleft zone where neural synapse carries out information through the pass of neurotransmitters from neuron to other. Mainly, neurotransmitters as electrically charged species that plays that role might be subject to electrical circuits that would be favorable or against synapse. Because this, the Boltzmann equation is employed as theoretical basis to explore electrodynamics of neurotransmitters. According to results of paper, circuit have been identified at cleft and the resultant electric devices have been corresponded to events of synapse. In particular, the resulting and dissipation power have shown certain correspondence to cellular behavior of neurotransmitters as to their role to complete the event of synapse. Thus, it would validate the collision's term of Boltzmann as to explain the apparition of inverse currents in cleft.

ABSTRACT. Mathematical models play a fundamental role in the treatment of Type 1 Diabetes Mellitus (T1DM), allowing to know the blood glucose dynamics with such accuracy that insulin therapy is significantly improved. The interpretability of these models is important for their clinical application and validity in research. In this work, the interpretability of different glucose model structures is evaluated by means of a qualitative and quantitative analysis, highlighting the importance of correlating model parameters with human physiology, thus strengthening their robustness and usefulness. This study allowed drawing relevant conclusions to improve the accuracy and applicability of mathematical models in the treatment of T1DM, moving towards more personalized and effective insulin therapies.

ABSTRACT. This article presents a studys, analysis of the Array Microstrip Antenna with circular polarization for WBAN/WIMAX applications. The analysis focuses on a single element and a two element microstrip patch, each designed at 2.5 GHz. Both antennas were crafted on a microstrip substrate, sporting a thickness of 1.6 mm and a dielectric constant of 4.3. An evaluation was conducted that contrasted the efficiency of the engineered antenna array with that of the standalone antenna, with particular emphasis on VSWR, gain, and directionality. Furthermore, the SAR level was scrutinized under flat conditions within the human tissue representation.

ABSTRACT. Cardiac MRI makes it possible to examine the morphology of the heart and blood vessels. The 4D Flow MRI sequence is used to diagnose complex cardiovascular diseases with high accuracy. Segmentation of the pulmonary arteries and aorta is essential to quantify various hemodynamic parameters. However, manual segmentation of these images presents significant challenges due to factors such as low signal-to-noise ratio and phase accumulation errors. To address this challenge, we propose a semantic segmentation cascade for automatic segmentation of pulmonary artery branches and aortic sections obtained from 4D Flow MRI angiographic images. The obtained results show that our methodology achieves accurate and consistent segmentation in segmenting artery bifurcations and sections in a limited dataset, which underlines its potential applicability in clinical settings.

ABSTRACT. This article presents a system for counting and analysing biological samples that includes a natural language interface for user interaction. This system uses artificial intelligence and big data for sample processing. Additionally, spatiotemporal samples are stored in the cloud, enabling automated classification and improved management.

ABSTRACT. This article examines the management of user traffic to the network access point and within the network, from the user’s access point to the destination server containing the required information. This study is conducted within the architecture of Software-Defined Networks (SDN). Two models are analyzed. The first model aims to minimize latency from the user to the network access point and within the network to retrieve information from the server while also minimizing the number of controllers. This model has a quadratic relationship between switches and controllers, which could affect performance. Subsequently, the second model linearizes the quadratic relationship. The goal is to manage the network by minimizing user access latency, considering that users will select the connection point with the minimum distance from the network access points to the destination server. For our research, we used 13 real Benchmark networks subject to extensive tests using our models and the computational power of the Gurobi software to find solutions. This work represents a pioneering effort in mathematical optimization modeling in this specific network paradigm. Detailed analysis of the test results shows that the models’ effectiveness is closely linked to the network’s scale. Additionally, compared to the first, the second model’s performance generally performs worse in larger networks. These insights offer a deep understanding of the models’ functionalities and optimization dynamics. They are finally, enhancing network efficiency and performance in SDNs.

ABSTRACT. This document details the implementation of software specialized in human face orientation on color images. To achieve this, a deep learning convolutional neural network called MobileNet-V2 has been used as a basis to create MobileNetV2_FO. To estimate the parameters of MobileNetV2_FO, transfer learning and fine-tuning techniques have been employed. Under this paradigm, prior knowledge has been leveraged, and this artificial intelligence has been specialized in face orientation classification. For training and inference, the PANDORA database has been utilized, which contains images of faces in nine orientations. The experimental results demostrate that MobileNetV2 FO achieves an accuracy of 94.095%, with precision and sensitivity both at 94.00%. Furthermore, in all cases, the value of the AUC (Area Under the Curve) exceeds 94.3%, with an average precision (AP) greater than 83%. This solution provides a competitive res- ponse in the identification of face orientation, rendering it an effective, efficient and versatile tool for a wide range of applications based on the human face, such such as security in biometric systems or driving assistance systems

ABSTRACT. Recent advances in wave energy exploitation highlight the potential of wave energy converters (WECs) to significantly contribute to global clean energy demands. Numerous concepts for ocean wave energy harvesting have been developed, with specific attention to internal reaction mass (IRM) systems, which are able to enclose the mechanical and electrical conversion stage within the floating hull itself, substantially reducing maintenance costs. Recent studies propose optimizing the performance of IRM systems by adjusting the behavior of the internal mass via optimal control procedures, significantly improving mechanical power available at the PTO. However, these studies often overlook energy losses associated with altering the internal mass dynamics itself, rendering any performance enhancement conclusions potentially overoptimistic. This paper addresses this gap by proposing a modified control objective that penalizes these losses, offering a balanced approach to optimizing the overall performance IRM WEC systems. The resulting optimal control problem is transcribed in a direct fashion, using moment-based theory, resulting in a finite-dimensional tractable nonlinear program, with application to a gyropendulum-based WEC. We show that the modified control objective is effectively able to trade-off the energy available at the PTO with the corresponding gyropendulum losses.

ABSTRACT. Lung cancer is the most common cause of mortality among various types of cancer globally, hence the pursuit of alternative methods for early identification of lung cancer should be a significant consideration, especially in low-resource settings. This work presents an alternative for lung cancer risk recognition based on machine learning. Utilizing lifestyle questionnaire data, we collected, analyzed, visualized, and processed data to train machine learning models. After balancing the data through SMOTE, our proposed model, XGBoost, was fine-tuned using the GridSearchCV method, to which we allocated 70% of our data, leaving the remainder for testing. The results obtained by the model for the first dataset included accuracy and F1 score of 96.50%, with precision and sensitivity of 96.51%. For the second dataset, an accuracy of 95.83%, F1 score of 95.83%, precision of 96.27%, and sensitivity of 95.83% were achieved. Additionally, through LIME, we obtained local interpretability of predictions, thus contributing to a more transparent understanding of model behavior and providing a clearer perspective on machine learning-based solutions.

ABSTRACT. The aim of this work is to remove systematic noise from transmission spectroscopy observations of exoplanets observed from the Earth’s surface using unsupervised clustering and multiresolution analysis (MRA) based on a Daubechies 12 as mother Wavelet. This will train a neural network of NARX topology (Nonlinear Autoregressive Model with Exogenous Inputs) which learns through the proposed method to remove the noise present in the observations. Within the method, it is proposed to remove the noise using the MRA in two different ways: applied for each wavelength observed to then recompose the Light Curve and directly on the Light Curve. From these, the best method is chosen based on the root mean square error (RMSE) metric, which was of the order of 10−7 for the best neural network in this research.

ABSTRACT. Deep learning techniques have achieved considerable success in pattern recognition and data classification tasks. However, existing Deep Neural Network (DNN) models are computationally expensive, limiting their effective utility in applications requiring performance guarantees in latency and throughput. The parallelism in DNN operations makes them ideal for modern GPU architectures. Nvidia’s TensorRT tool aids in mapping neural network-based algorithms to Nvidia GPUs, promising resource optimization and inference acceleration. Given the closed nature of TensorRT's internal details, its effectiveness can only be evaluated empirically. With the rapid evolution of GPU hardware and software, periodic evaluations are necessary to validate findings, develop efficient methodologies, and verify the generality of techniques across different problems. This paper reports results obtained from a systematic experimental evaluation of TensorRT's optimization capabilities on GPUs of various ranges, from embedded systems to desktop computers, providing quantitative data on its advantages and limitations. The findings offer guidelines for optimizing customized algorithms in scientific and industrial applications beyond standardized benchmarking examples.

ABSTRACT. Cultural heritage represents one of the areas with a growing interest in applying new technologies because of the wide variety of offered content. In particular, a significant application for new technologies involves enhancing the user's cultural experience by providing personalized content selected according to the user's preferences. For this purpose, one of the most commonly exploited tools is Recommender Systems, which can filter and analyze data to support users according to their preferences. Moreover, in the personalization phase, the application to cultural heritage requires the consideration of the environmental situation experienced by the user. Therefore, Recommender Systems evolve into Context-Aware Recommender Systems integrating data acquired through sensors. As a result, the objective of this work consists of describing a four-layer architecture for improving cultural experience through the Internet of Things and Context-Aware Recommender Systems. In particular, the proposed recommendation approach is tested on a dataset in the literature, yielding promising results in accuracy.

ABSTRACT. In recent years, the electric vehicle (EV) industry has experienced significant advancements, simultaneously driving substantial progress in battery technology. The evolution of battery systems necessitates enhancements in charging infrastruc- ture to attain elevated power levels during the charging process, thereby minimizing charging time. Various algorithms have been developed for driving battery charging; however, these algorithms necessitate the creation of diverse controllers to generate precise trigger signals for the semiconductors within the various power converters utilized in charging stations. This work presents the design of an innovative model-free control system for Type I impedance network partial power converter (PPC) in which a Deep Reinforcement Learning (DRL) agent generates control signals during the different charging stages. Particularly, a Twin- Delayed Deep Deterministic Policy Gradient (TD3) algorithm is used to substitute the inner control loop of traditional control systems. To this end, different agents were designed, trained, and tested inside a built simulation environment. It is worth noting that TD3-based control allows for the optimal functionality of a type I impedance network PPC within the context of EV battery charging applications, according to the specified CC-CV charging algorithm. Empirical results revealed that the battery system reached an 80% state of charge in under 8 minutes starting from an initial 20%.

ABSTRACT. This paper presents a proposition for a configuration designed for an open-winding permanent magnet synchronous motor featuring two distinct DC link voltages. The optimization focuses on the ratio between these two unequal DC link voltages. A two-level voltage source inverter supplies each side of the open-winding permanent magnet motor. By leveraging isolated DC voltages, a maximum of 49 different voltage vectors can be applied to govern the motor operation. A model predictive control approach is utilized to ascertain the most suitable voltage vector for the given switching period. The method aims to enhance performance by minimizing harmonic distortion through optimizing voltage ratio. Simulation outcomes demonstrate that the refined ratio of DC link voltages reduces the total harmonic distortion factor of stator current compared to alternative ratios.

ABSTRACT. This work deals with advances in the implementation and use of the Controller Area Network to communicate power converters distributed in a Microgrid in Universidad del Bío-Bío, Chile. The objective is to allow the future implementation of a Secondary Control and Tertiary Control, expanding the capability of the Laboratorio de Acondicionamiento y Conversión de Energía of Universidad del Bío-Bío to test microgrid control strategies. The manuscript reviews the Microgrid implemented in Universidad del Bío-Bío, the implementation of the CAN bus among the different devices in the microgrid, focusing on the digital control platform used in the power converters, and a preliminary test to set the power converter references using the CAN bus.

ABSTRACT. This document proposes the use of repetitive control technique to be implemented in a three-phase inverter, considering its connection to a microgrid. The goal is to control the system current while simultaneously implementing harmonic compensation to current. The Park and Clark transforms will be used to simplify the analysis and control of three-phase systems, allowing them to work with more manageable coordinates. The effectiveness of harmonic compensation is confirmed through the DFT along with the results of the simulations in Matlab

ABSTRACT. Electric machines are one of the most important pieces of equipment in modern industrial processes. However, due to different applications and for energy efficiency, most of the electric machines are not directly connected to the power grid, but through a power converter. This allows to control the speed, torque, and therefore the currents and voltages, increasing the number of applications for higher power processing. This study compares three control strategies for induction motors: Indirect Field Oriented Control (IFOC), Direct Torque Control (DTC), and Predictive Torque Control (PTC). Their performance is evaluated in terms of efficiency, steady-state and transient response. Simulation results indicate that IFOC achieves the lowest total harmonic distortion (THD), while PTC shows promise as an alternative for power electronics applications. The results provided in this investigation demonstrate the advantages and disadvantages of the techniques proposed above

ABSTRACT. Generating common-mode voltage (CMV) by an inverter is a significant concern. In photovoltaic systems without transformers, fluctuations in CMV lead to a flow of leakage current and high-frequency harmonics. Using a limited number of switching states may reduce the leakage current. The cost of this act is the deterioration of the current quality. Model predictive control, reinforced by the virtual voltage concept, is useful for compensating for this drawback without needing system linearization. However, the computational burden of the processor will increase. To overcome this issue, this research proposes a remote state deadbeat technique. A cost function is defined to choose the optimal voltage vector from among ten candidate voltage vectors. This method reduces both line current distortion and the leakage current. Also, the computational burden is almost similar to that of the remote state predictive method.

ABSTRACT. This panel will delve into the latest advancements in agricultural digitalization, focusing on IoT, AI, and advanced data platforms. We will explore how these technologies are optimizing productivity, resource efficiency, and environmental impact, addressing challenges such as economic barriers and integration with traditional methods. Additionally, we will discuss the vital role of collaboration in driving sustainable and scalable solutions for climate change and food security.

ABSTRACT. In agriculture, precise and timely meteorological data is critical for optimizing crop management, improving yield quality, and mitigating the impacts of adverse weather conditions. However, existing meteorological services often limit data downloads to three continuous months, constraining long-term analysis essential for agricultural purposes. This study introduces an advanced R Shiny application designed to automate the retrieval and analysis of real-time meteorological data from the Dirección General de Aeronáutica Civil and the Dirección Meteorológica de Chile - Servicios Climáticos. The application provides an intuitive graphical user interface that allows farmers, researchers, and agricultural advisors to access and interact with over 15 climatic parameters, such as temperature, humidity, wind direction, and precipitation. The application supports downloading data for selected periods and visualizing the area of influence for each station to select the most relevant data. By integrating real-time data streaming, this tool enhances decision-making processes, supports efficient resource management, and fosters the adoption of digital agriculture practices. The study demonstrates the application's efficacy through a case study in Chile, highlighting its potential to improve water management, pest control, and overall crop resilience. Integrating cutting-edge data retrieval technologies in agriculture empowers stakeholders with actionable insights and promotes sustainable agricultural practices in response to climate variability.

ABSTRACT. This research introduces a digitised mathematical model that enables exploration and determination of the most relevant indicators for optimal rootstock selection in Chilean viticulture. Through this computational simulation, an effective tool is aimed to be provided for enhancing decision-making and fostering the production of resilient, high-quality vines in response to changing climatic conditions. The results obtained encourage the scientific community and farmers to utilise this tool for improving the interaction and selection of rootstocks in the wine sector, which is pivotal for Chile's social and economic development. Rootstocks emerge as a valuable alternative to confront climate change and enhance vine adaptability to varying environmental conditions. Biomathematical modelling stands out as an essential tool for investigating the intricate interaction between plants and soil, particularly in the context of rootstocks in viticulture. Moreover, the incorporation of Graphical User Interfaces (GUIs) within the MATLAB software paves a way to bridge scientific knowledge with field farmers, enabling a more informed and efficient decision-making in the selection of optimal rootstocks for wine production in a dynamic environment.

ABSTRACT. Accurate and spatially reliable temperature data are crucial for effective agricultural management and climate adaptation. This study presents an innovative methodology for temperature correction using topographic factors and adiabatic models. Initial temperature information, sourced from the automatic weather stations (AWSs) of the Dirección General De Aeronáutica Civil (DGAC), often has limited spatial representation, covering approximately 30 kilometers in radius, however, significant temperature variations within this area require refined data for precise agronomic applications. This research proposes a method to spatialize and correct temperature data to enhance its reliability for agronomic use. More accurate spatial temperature maps are achieved by utilizing digital elevation models (DEMs) and applying corrections restricted to locations within 100 meters of elevation difference from the AWSs. These corrections include wet and dry adiabatic lapse rates, which are applied appropriately. The corrected temperature data are then used to create detailed spatial maps, essential for modeling agronomic variables in changing environments. These maps enable better decision-making for irrigation scheduling, crop stress monitoring, and other critical agricultural practices. The methodology was tested in four distinct sites within the central-southern macrozone of Chile, specifically in the regions of O'Higgins, Maule, Ñuble, and Biobío. The results from these case studies provided valuable insights into the applicability and accuracy of this technique in real-world agricultural settings, demonstrating significant improvements in climate resilience and sustainability. By improving the spatial accuracy of temperature data, this methodology supports more effective resource management and enhances the sustainability of agricultural operations. The findings highlight the potential of advanced topographic and adiabatic corrections to transform temperature data analysis, providing a valuable tool for climate-adaptive agriculture.

ABSTRACT. The paradigm of intelligent agriculture has significantly impacted the way in which crops are cared for and maintained. As an alternative, the use of unmanned aerial vehicles (UAVs), or drones, for the purposes of fertilizer application and the prevention of crop pest spread has been proposed. We put forth a solution based on UAVs equipped with computer vision algorithms for the early detection of plants infested with aleurothrixus floccosus in citrus crops. The drone is equipped with a high-definition camera that records the affected plantation, and this video is shared via streaming with a personal computer (PC) that executes the computer vision algorithm to determine the presence or absence of the pest. We selected the open-source real-time object detection and image segmentation model, You Only Look Once (YOLO) version 10s, as a preliminary approach. The numerical results demonstrate that the average precision is 73% during the training stage and reaches up to 76% during the operation stage.

ABSTRACT. In this work, an electronic probe prototype for a water quality monitoring system is presented. This prototype can measure different parameters of water as the electrical conductivity, temperature, potential of hydrogen (pH), dissolved oxygen, and total dissolved solids. The measurements carried out with the homemade multiparametric probe were compared with those made by commercial and professional probes, showing similar precision in the data acquisition. The prototype was built with low-cost sensors for in-situ measurements, and they are integrated with the Internet of Things (IoTs) to visualize and send the information in real-time. Also, the interface is quite friendly with the intention that communities can manipulate the portable systems to make citizen water quality monitoring.

ABSTRACT. Accurately estimating chill hours and growing degree days is crucial for monitoring and predicting plant development, particularly under climate change and its impacts on agriculture. This paper comparatively evaluates four methods for calculating chill hours and growing degree days: the simple hour method, the simple triangle method, the double hour method, and a proposed combined method. A qualitative and quantitative comparison is presented. The proposed combined method integrates aspects of the simple hour, simple triangle, and double hour approaches to obtain more accurate estimations. Quantitative examples are provided to illustrate the differences in estimations among the methods. This comparative analysis aims to guide researchers and practitioners in selecting the most appropriate method for their specific crop, region, and climate conditions, contributing to improved agricultural management and decision-making.

ABSTRACT. The water crisis is one of the most serious problems for agriculture worldwide, with the availability of water quantity and quality being one of the sustainable development objectives. In Chile, one of the large industries is wine, which has also been affected by the drought due to the lack of water available for irrigation. To address this situation, innovate alternatives are sought that are viable to apply in various contexts, which is why the case of building and implementing a fog catcher in the Casablanca valley, Valparaiso, Chile, with monitoring of environmental and water parameters, is presented. Collected, through a low-cost system implemented with the ESP8266 Node MCU board. Continuous monitoring allows you to periodically study environmental conditions, such as having the real-time availability of the water resource to be able to make decisions regarding efficient use.

ABSTRACT. In power grids, Supervisory Control and Data Acquisition (SCADA) systems are essential for collecting data from the diverse network elements, including electrical measurements and equipment status among others. In the last decades, power grids became increasingly complex. For example, the so-called microgrids span from the electricity generation to consumption, close to the end-users. These systems can operate independently from the grid or remain grid-connected. Geographically concentrated microgrids interconnect diverse energy resources, like photovoltaic panels, batteries, and diesel generators, while adopt new technologies for continuous evolution. This paper proposes a design for adapting SCADA systems to microgrids, with a focus on achieving flexibility and scalability within small and heterogeneous networks of Intelligent Electronic Devices (IEDs), the so-called MicroSCADA. The proposed design structure is shown, outlining the key components required for system-agnostic implementations. Finally, the design is tested and validated in an experimental setup, with a set of virtualized IEDs. And shows a promising scheme for a real industrial flexible application, independent from the type of IED.

ABSTRACT. Medium voltage circuit breakers are critical elements in properly functioning the electrical system. Disruptive discharges occur when the current completely crosses the insulator, separating two conductors with different potentials. Numerous studies have been carried out on disruptive discharges. However, there are still challenges in understanding the factors that influence it and improving its prediction and control, especially for students studying careers related to this subject. In the present work, the mathematical simulation of the electrical breakdown of oil in a medium voltage switchgear is carried out using commercial software to improve the understanding of this phenomenon. The results obtained are compared with the experimental part implemented in the high voltage laboratory of the manufacturer TERCO belonging to the Universidad Politécnica Salesiana , which consisted of obtaining values of breakdown voltages using oil as a dielectric medium in a 13.8 kV capacitor switch with which the closing and opening of its contacts were performed.In these results it was observed that the voltage rises at the terminals of the power switch, in the opening of the switch with inductive resistive load, the circuit to be fed with 220 volts and let it reach steady state, the voltage at the terminals of the switch had a voltage of zero volts, but after opening the circuit step to 280 volts.

ABSTRACT. The construction sector is responsible for emitting 30\% of worldwide greenhouse emissions. Sustainable construction addresses this problem by designing infrastructures under the energy-efficient concept. This paper presents an energy-efficient optimal methodology for designing the thermal envelope and energy support technologies of the infrastructure, according to the energy qualification established in 21.305 law. The methodology consists of carrying out a mathematical formulation of the optimization problem for an energy solution implementing a home energy rating (CEV) as an input parameter, which allows obtaining specific results for all the elements that participate in the property's requirement. Specifically, simulations were carried out of a house located in the north and south of Chile considering different parameters such as floor, walls, windows, etc. Simulating in each area and restricting the energy rating where the sensitivity of the problem to the different parameters is observed. Finally it is possible to conclude about the minimum energy requirements.

ABSTRACT. The Energy Management Portal (PGEN) represents a significant effort to optimize electrical energy consumption in 670 units across Brazil’s Federal Network of Professional, Scientific and Technological Education (EPCT Network). Funded by the National Electric Energy Conservation Program (PROCEL), PGEN aims to gather consumption data in real-time and publicly, allowing accurate diagnoses and assertive actions to reduce consumption. Furthermore, this project involves students, promoting awareness about the rational use of energy and training them for future energy management practices. In short, PGEN is a strategic initiative that combines technology, transparency, and education, aiming not only at financial savings but also at the formation of a sustainable and responsible culture in relation to electrical energy in educational institutions.

ABSTRACT. The electrical grid is dealing with several new factors and agents, such as intermittent renewable energy and electrical vehicle, which compromise the good performance of the grid. In order to ensure a suitable electrical supply despite of these new factors, the need of an smart grid configuration evident. Moreover, the integration of new agents on the grid sometimes leads to the occurrence of electrical faults. The detection and classification of these electrical faults is of vital importance to guarantee the stability and good quality of the electrical supply. Nowadays data-driven methods based on machine learning techniques for detection and classification of electrical faults are gaining interest from the scientific community thanks to the good results they offer. In this paper, a solution for the detection and classification of electrical faults is presented based of machine learning algorithms and mainly driven in five main steps. The good results prove the good performance of the proposed solution.

ABSTRACT. The Galapagos Islands, a fragile and unique ecosystem considered a living laboratory, are threatened by climate change and heavy dependence on fossil fuels for energy and transportation. This study examines the accelerated energy transition process needed to decarbonize the islands by 2040, with an emphasis on transportation. Scenarios are developed using the EnergyPLAN tool, identifying energy potentials, benefits of renewable energy, and barriers to overcome. Results indicate at least 40GWh of transportation energy will be needed by 2040, with total energy mix exceeding 100GWh. Accelerated adoption may require policies such as carbon taxes and emissions trading. Electromobility powered by renewables will be key for transportation decarbonization. Concrete near-term actions are essential to enable the transition and avoid future energy scarcity. Policies to facilitate implementation, backed by sufficient budgets, place great responsibility on decision makers to safeguard this treasure for humanity.

ABSTRACT. The ongoing transition of electric grids has posed significant operational challenges for System Operators (SOs), which has led to the implementation of new data-based techniques and analysis. Various system operators and research projects have started to share different datasets, especially frequency data. This work proposes the development of a system based on LSTM networks for the detection of contingencies and the identification of characteristic patterns in frequency signals, while only storing relevant information. Therefore, it showcases a new tool that contains actual classified frequency events obtained from phasor measurement unit (PMU) recordings in the Chilean electric system.