The Isolation Forest (IF) algorithm is effective in detecting anomalies in critical infrastructure, but its performance depends on the proper setting of five hyperparameters: sample size, number of trees, maximum tree depth, maximum number of features and detection threshold. Static tuning of these parameters is inefficient and poorly adaptable to dynamic environments. This paper proposes a multivariate autotuning method that automatically optimises these hyperparameters by: (1) adaptive adjustment of the sample size based on the standard deviation of the anomaly scores, (2) selection of the number of trees according to F1-score stabilisation, (3) control of the maximum depth based on the average isolation rate, (4) adjustment of the maximum number of features according to the variance of the data, and (5) optimisation of the detection threshold by minimisation of a cost function. The auto-tuning procedure has been validated in the detection of anomalies in drinking water networks, s howing an F1-score improvement of 7.5% and a reduction of the execution time by 22.55% compared to static configurations, demonstrating its feasibility for real-time systems.

FOG computing is an essential paradigm for Internet of Things (IoT) environments, as it enables the decentralisation of data processing to make better use of resources. The rapid proliferation of IoT devices has created new challenges for this middle layer, especially in terms of reconfiguration and dynamic scalability, processes that have traditionally been managed manually. This study presents an architecture for the FOG layer that implements dynamic control and scaling automation through an advanced control layer composed of a coordinator that calculates the optimal distribution of computational load and a balancer that manages traffic routing. The central component of the proposal is the Simple FOG Management Protocol (SFMP), designed specifically to facilitate automated resource management, ensuring efficient allocation of computational capacity, fair load balancing, and robust fault tolerance mechanisms. Experimental validation in several complex IoT environments at the University of Alicante demonstrated a substantial optimisation in resource utilisation with a reduction in energy consumption of between 40 % and 65 %. All this confirms the flexibility and adaptability of the system to dynamic variations in computational demand in large-scale IoT implementations and its benefits over traditional proposals.

AutoForest es una aplicación de escritorio para la detección de anomalías en datasets de infraestructuras críticas utilizando el algoritmo Isolation Forest con auto-ajuste automático de hiperparámetros. La aplicación funciona aportando un dataset, y ejecutando una serie de algoritmos que permiten el cálculo optimizado de los hiperparámetros que actúan dentro del algoritmo Isolation Forest. Se obtiene un ajuste independiente del conocimiento del usuario.

Optimising energy consumption has become a priority objective for companies and organisations, especially in essential services such as drinking water supply. Traditional strategies of maximising water storage generate energy inefficiency by forcing pumps to move large masses of water. These strategies also generate other negative effects, such as water loss in case of incidents and stress on aquifers. The problem addressed is the need to optimise the operation of drinking water production plants, seeking a balance between minimising stored water, guaranteeing supply and reducing energy consumption. Traditional optimisation systems have limitations in the number of variables they can handle and in the search for solutions in large search spaces. The optimisation problem is the search for solutions in an immense and computationally intractable data space. In this work, a new algorithm based on predictive artificial intelligence and the search for optimal values in distributed and parallel data spaces is proposed. This algorithm includes a consumption prediction module and a control module that calculates the optimal on/off sequence of the pumps. The algorithm uses a heuristic approach to define the search space and can be distributed and parallelised for computational feasibility. The prediction is based on a KNN algorithm that weights historical consumption, giving more weight to more recent data. To address the problem of large search spaces, the algorithm has been parallelised by distributing the computation of the optimisation matrix over several nodes. This allows the computation of local optima and the subsequent union of the results in a central node to obtain the global optimum. The proposed solution has been validated using real water infrastructure data in a city of approximately 5000 inhabitants in southeastern Spain. The validation shows that the new algorithm is more efficient than previous operational strategies, guaranteeing the service with a lower energy consumption of approximately 6%. The implementation of the algorithm has made it possible to maintain the water level

Fog computing can provide an effective solution to the challenges presented by today’s ever-emerging Internet of Things (IoT) infrastructures. As the number of interconnected devices progressively increases, these infrastructures require better solutions to ensure high scalability and processing capacity, along with an efficient use of available resources. This is why this paper presents a distributed Fog architecture, specifically designed to address the challenges and difficulties presented by heterogeneous IoT environments. This Fog architecture is used as an intermediate layer between the IoT devices and the final layer, it has been designed after the previous analysis of the requirements to be met for the solution, then the modularization of the architecture has been carried out so that it can be easily distributed, and finally, an implementation has been generated on a real environment as a validation case of the proposal.

Optimising energy consumption has become a priority objective for companies and organisations, especially in essential services such as drinking water supply. Traditional strategies of maximising water storage generate energy inefficiency by forcing pumps to move large masses of water. These strategies also generate other negative effects, such as water loss in case of incidents and stress on aquifers. The problem addressed is the need to optimise the operation of drinking water production plants, seeking a balance between minimising stored water, guaranteeing supply and reducing energy consumption. Traditional optimisation systems have limitations in the number of variables they can handle and in the search for solutions in large search spaces. The optimisation problem is the search for solutions in an immense and computationally intractable data space. In this work, a new algorithm based on predictive artificial intelligence and the search for optimal values in distributed and parallel data spaces is proposed.

Herramienta de generación de datasets sintéticos para entornos IoT. El objetivo del software es proporcionar a investigadores una herramienta simplificada y eficiente para la generación de conjuntos de datos sintéticos basados en patrones personalizados de comportamiento de dispositivos IoT. La aplicación permite modelar comportamientos lo más realistas posible y controlados de sensores para su uso en el desarrollo y validación de proyectos de I+D+i. La herramienta permite la definición estructurada de los datos emitidos por un dispositivo de forma que puede replicarse cualquier fuente de datos.

The optimised operation of drinking water supply infrastructures is a crucial objective for companies and cities, facing the complexity of multiple conflicting objectives and factors such as climate change and demand fluctuations. This paper presents an intelligent and predictive control system designed for the efficient and safe management of these resources. The proposed system combines a multi-agent model with Artificial Intelligence algorithms. It uses the KNN algorithm for consumption prediction, highlighting its high accuracy (average relative error of 6%) and its ability to provide interpretable results that reflect human expert knowledge, avoiding the “black box” behaviour of other AIs. An essential component is the simulation tool developed. This allows the configuration, testing, validation and adjustment of the control system and its hyperparameters in a safe environment, preventing damage to critical infrastructure and optimising investments. The solution has been successfully validated in two real drinking water infrastructures in south-eastern Spain (serving 5,000 and 90,000-200,000 inhabitants). The results demonstrate a tangible improvement in pressure stabilisation, maintenance of optimal water levels (reducing unnecessary storage and the formation of by-products) and minimisation of pump starts/stops, which translates into energy savings and a longer service life for the equipment. In addition, this approach is adaptable to various infrastructure configurations.

Este software permite la simulación de una infraestructura de generación y abastecimiento de agua potable a través de la gestión utilizando sistemas multi agente. Este modelo de gestión permite abordar la optimización multi objetivo a través de software distribuido. La herramientas permite la parametrización de un red de abastecimiento de agua y de los sistemas de producción mediante la extracción de pozos de agua potable. El simulador muestra la evolución de la infraestructura en el tiempo y permite encontrar los parámetros más adecuados para la configuración del sistema de control.

To train LLM using in-context learning techniques, it is necessary to provide it with knowledge in a way that is understandable to the model. To do this, it is provided with a prompt indicating the level of specialization along with certain guidelines and requirements that the LLM must take into account when reasoning, and a dataset with the knowledge that the LLM must learn, which in this case is formalized through a series of questions and answers about the work environment. The results generated by the LLM after training with the prompt and the dataset are also collected.

Optimising energy consumption has become a priority objective for companies and organisations, especially in essential services such as drinking water supply. Traditional strategies of maximising water storage generate energy inefficiency by forcing pumps to move large masses of water. These strategies also generate other negative effects, such as water loss in case of incidents and stress on aquifers. The problem addressed is the need to optimise the operation of drinking water production plants, seeking a balance between minimising stored water, guaranteeing supply and reducing energy consumption. Traditional optimisation systems have limitations in the number of variables they can handle and in the search for solutions in large search spaces. The optimisation problem is the search for solutions in an immense and computationally intractable data space. In this work, a new algorithm based on predictive artificial intelligence and the search for optimal values in distributed and parallel data spaces is proposed.

The application applies the Isolation Forest algorithm to multivariate time-series data uploaded as a CSV file. It scores each data point with an anomaly probability, classifies points above a configurable threshold as anomalies, groups consecutive anomalies into incidents, and provides a set of interactive charts to explore the results.

This paper presents the Data Quality Assurance Model (DQAM), a formal model and multilayered architecture designed to guarantee data integrity and robustness in Reliable and Quality-Assured Information Systems. Recognising that inaccurate or corrupted sensor data can lead to system collapses and false alarms in critical services, the DQAM provides a standardised and systematic flow of actions to ensure data excellence for Artificial Intelligence (AI). The architecture is structured into three specialised layers (Acquisition, Processing, and AI Adequacy), implementing formal transformation functions that act as a rigorous filter against data degradation. A core contribution is the mapping of these functions to ISO/IEC 25012 and 5259-2 standards, providing a practical framework for reliable information management. It should be noted that quality dimensions regarding timeliness and data volume are outside the scope of this work, as they depend on external data issuers and end-service requirements. The model’s viability is validated through a real-world implementation on a university campus managing millions of data points, demonstrating its capability to optimise performance—achieving a speedup of up to 43%—and prevent service malfunctions. This work bridges the gap between raw IoT streams, and the high-integrity standards required by modern AI-driven applications.