Óscar García

Video violence detection using artificial intelligence plays a key role in public safety applications. Although convolutional and recurrent neural networks are widely adopted for this task, the actual contribution of temporal modeling over strong frame-level representations remains insufficiently analyzed. This work provides a systematic study of video violence detection models under a unified experimental framework. We investigate whether violence can be reliably detected from individual frames without explicit temporal modeling, evaluate the effectiveness of combining CNNs with LSTM and Bi-LSTM layers, and analyze the impact of architectural and hyperparameter choices, including neuron configuration and backbone selection (VGG-16 vs. VGG-19). Experiments are conducted on three widely used benchmark datasets. Our results show that frame-level analysis using a pre-trained VGG-19 network, combined with a simple aggregation strategy, achieves competitive performance, reaching 95% accuracy on Hockey Fights and 96% on Violent Flow. While Bi-LSTM layers can provide moderate improvements of up to 4% over standard LSTM models in certain datasets, these gains are not consistent across all scenarios. Furthermore, variations in hyperparameter configurations do not systematically lead to improved performance. Overall, this study highlights that increased architectural complexity does not always translate into better results and that, in several cases, simple frame-based approaches can rival more complex temporal models. These findings provide practical insights into the cost–benefit trade-off of temporal modeling for video-based violence detection.

Construction planning is a critical and complex phase in the deployment of large‐scale renewable energy infrastructure. This study applies artificial intelligence techniques to a domain‐specific problem that has traditionally relied on expert judgement: the generation of detailed construction schedules for photovoltaic power plants. As renewable generation is a key part to meet the challenges of energy transition, the implementation of large projects has increased in recent years and this trend is expected to continue in the future. The main difficulty in meeting construction deadlines is the elaboration of an adequate planning. A tool that automatically generates schedules can be of great help to set up an initial baseline planning. To this end, this work compares five artificial intelligence techniques, on a data set consisting of real examples of successfully completed projects. The evaluation of the results obtained on test data shows that Adaptive Neuro‐Fuzzy Inference System (ANFIS) is the technique that obtains the best performance in all error metrics, although it entails a high computational cost. The model thus obtained manages to generate a complete construction schedule with an error of 8% of the total duration. The use of metrics as MAE, MSE and provides a robust understanding of prediction accuracy, variability, and fit. These metrics are commonly used in project planning evaluations and help interpret model behaviour under different error profiles. Additionally, the resulting 8% total duration error implies a deviation of around 24 days in a 300‐day project, which is highly actionable in real‐world solar project management. The findings not only demonstrate the feasibility of using AI for solar construction planning, but also lay the groundwork for the development of intelligent software tools or platforms that could support planners in the renewable energy sector. While this study focuses on photovoltaic plants, the approach is extendable to other power plants as wind farms, combined‐cycle or nuclear plants, or even to other construction projects.

The effects of climate change and the rapid growth of societies often lead to water scarcity and inadequate water quality, resulting in a significant number of diseases. The digitalization of infrastructure and the use of Digital Twins are presented as alternatives for optimizing resources and the necessary infrastructure in the water cycle. This paper presents a framework for the development of a Digital Twin platform for a wastewater treatment plant, based on a microservices architecture which optimized its design for edge computing implementation. The platform aims to optimize the operation and maintenance processes of the plant’s systems, by employing machine learning techniques, process modeling and simulation, as well as leveraging the information contained in BIM models to support decision-making.

Yearly population growth will lead to a significant increase in agricultural production in the coming years. Twenty-first century agricultural producers will be facing the challenge of achieving food security and efficiency. This must be achieved while ensuring sustainable agricultural systems and overcoming the problems posed by climate change, depletion of water resources, and the potential for increased erosion and loss of productivity due to extreme weather conditions. Those environmental consequences will directly affect the price setting process. In view of the price oscillations and the lack of transparent information for buyers, a multi-agent system (MAS) is presented in this article. It supports the making of decisions in the purchase of sustainable agricultural products. The proposed MAS consists of a system that supports decision-making when choosing a supplier on the basis of certain preference-based parameters aimed at measuring the sustainability of a supplier and a deep Q-learning agent for agricultural future market price forecast. Therefore, different agri-environmental indicators (AEIs) have been considered, as well as the use of edge computing technologies to reduce costs of data transfer to the cloud. The presented MAS combines price setting optimizations and user preferences in regards to accessing, filtering, and integrating information. The agents filter and fuse information relevant to a user according to supplier attributes and a dynamic environment. The results presented in this paper allow a user to choose the supplier that best suits their preferences as well as to gain insight on agricultural future markets price oscillations through a deep Q-learning agent.

Context: Smart Energy is a disruptive concept that has led to the emergence of new energy policies, technology projects, and business models. The development of those models is driven by world capitals, companies, and universities. Their purpose is to make the electric power system more efficient through distributed energy generation/storage, smart meter installation, or reduction of consumption/implementation costs. This work approaches Smart Energy as a paradigm that is concerned with systemic strategies involving the implementation of innovative technological developments in energy systems. However, many of the challenges encountered under this paradigm are yet to be overcome, such as the effective integration of solutions within Smart Energy systems. Edge Computing is included in this new technology group. Objective: To investigate developments that involve the use of Edge Computing and that provide solutions to Smart Energy problems. The research work will be developed using the methodology of systematic mapping of literature, following the guidelines established by Kitchenham and Petersen that facilitate the identification of studies published on the subject. Results: Inclusion and exclusion criteria have been applied to identify the relevant articles. We selected 80 papers that were classified according to the type of publication (journal, conferences, or book chapter), type of research (conceptual, experience, or validation), type of activity (implement, validate, analyze) and asset (architecture, framework, method, or models). Conclusion: A complete review has been conducted of the 80 articles that were closely related to the questions posed in this research. To reach the goal of building Edge Computing architectures for Smart Energy environments, several lines of research have been defined. In the future, such architectures will overcome current problems, becoming highly energy-efficient, cost-effective, and capacitated to process and respond in real-time.

The Internet of Things (IoT) has become one of the most widely research paradigms, having received much attention from the research community in the last few years. IoT is the paradigm that creates an internet-connected world, where all the everyday objects capture data from our environment and adapt it to our needs. However, the implementation of IoT is a challenging task and all the implementation scenarios require the use of different technologies and the emergence of new ones, such as Edge Computing (EC). EC allows for more secure and efficient data processing in real time, achieving better performance and results. Energy efficiency is one of the most interesting IoT scenarios. In this scenario sensors, actuators and smart devices interact to generate a large volume of data associated with energy consumption. This work proposes the use of an Edge-IoT platform and a Social Computing framework to build a system aimed to smart energy efficiency in a public building scenario. The system has been evaluated in a public building and the results make evident the notable benefits that come from applying Edge Computing to both energy efficiency scenarios and the framework itself. Those benefits included reduced data transfer from the IoT-Edge to the Cloud and reduced Cloud, computing and network resource costs.