Industry 5.0 is driving a new era in industrial automation, where the collaboration between artificial intelligence (AI) and human supervision enables the development of smarter, more adaptive, and more efficient systems. Robotic trajectory generation is a clear example of this new paradigm. Metaheuristic techniques help automatically generate optimized trajectories, thereby improving operational efficiency. However, optimizing trajectories using AI alone also presents limitations. Starting from random trajectories, the optimization process becomes computationally expensive, especially in complex environments. In this context, initial input from human experts plays a crucial role: expert-defined trajectories provide structured, feasible, and contextual starting points that guide AI more effectively toward high-quality solutions. Therefore, this work proposes a novel human-guided trajectory optimization algorithm. In this way, human knowledge, Particle Swarm Optimization (PSO), and Pattern Search (PS) are efficiently combined. The results demonstrate that this approach significantly improves robotic system performance, achieving cycle time reductions of up to 16.69% compared to expert-defined trajectories. This approach establishes a solid framework for intelligent automation in Industry 5.0, promoting the development of more efficient, sustainable, and adaptive robotic systems.

The rapid expansion of wind energy has increased the need for reliable, data-driven condition monitoring to reduce operation and maintenance costs and improve turbine availability. This paper presents an anomaly detection framework that combines (i) Fast Fourier Transform (FFT) feature extraction to capture discriminative frequency-domain signatures, (ii) a graph-based formulation of multi-sensor dependencies learned with a hybrid Graph Neural Network (GraphSAGE + GATConv), and (iii) a genetic algorithm for automated hyperparameter optimization. Experiments are conducted on the publicly available Gearbox Reliability Collaborative (GRC) dataset from the National Renewable Energy Laboratory (NREL), using gearbox vibration signals from five selected accelerometer channels. Under a window-level protocol with a class-balanced 70/30 train–test split, the proposed model achieves training accuracy and test accuracy. On the held-out test set, it attains a macro-averaged precision of , recall of , and F1-score of , while detecting the Damaged condition with precision and recall. These results indicate a small generalization gap and outperform previously reported AI-based methods on the same dataset, supporting the effectiveness of the proposed GNN-driven approach for robust wind-turbine gearbox anomaly detection.

ABSTRACT In the context of industrial automation, optimising automated guided vehicle (AGV) trajectories is crucial for enhancing operational efficiency and safety. They must travel in crowded work areas and cross narrow corridors with strict safety and time requirements. Bio‐inspired optimization algorithms have emerged as a promising approach to deal with complex optimization scenarios. Thus, this paper explores the ability of three novel bio‐inspired algorithms: the Bat Algorithm (BA), the Whale Optimization Algorithm (WOA) and the Gazelle Optimization Algorithm (GOA); to optimise the AGV path planning in complex environments. To do it, a new optimization strategy is described: the AGV trajectory is based on clothoid curves and a specialised piece‐wise fitness function which prioritises safety and efficiency is designed. Simulation experiments were conducted across different occupancy maps to evaluate the performance of each algorithm. WOA demonstrates faster optimization providing suitable safety solutions 4 times faster than GOA. Meanwhile, GOA gives solutions with better safety metrics but demands more computational time. The study highlights the potential of bio‐inspired approaches for AGV trajectory optimisation and suggests avenues for future research, including hybrid algorithm development.

During the production of an animated film, professionals at the animation studio prepare thousands of notes. These notes describe improvements and corrections identified by supervisors and directors during daily meetings where the film’s progress is reviewed. After each meeting, these notes are manually distributed to the appropriate departments that need to address them. Due to the manual nature of this process, many notes are not assigned correctly, and the identified issues are not addressed, reducing the final quality of the film. This article describes and compares several approaches to automatically distribute notes using multi-label text classification with different language models (LM). Implemented methods include logistic regression models, encoder-only models such as the BERT family, and decoder-only models such as Llama 2 including fine-tuning and QLoRA techniques. Training and inference were conducted on a local RTX-3090. The results of the different techniques have been compared, achieving a maximum average accuracy of 0.83 and an f1-score of 0.89 with the fine-tuned Multilingual BERT model. This demonstrates the validity of these models for multi-label text classification, as well as their usefulness in a hitherto unexplored area such as animation studios. • Encoder-and decoder-only language models for multi-label text classification. • Large Language Models for automatic distribution of text notes into departments. • Transfer learning, in-context learning, fine-tuning large language models. • Real dataset of animated movie production studio text notes.

This paper presents an offline optimization method designed for use with industrial robots in environments with static obstacles. It is particularly useful in industry where stability and predictability are crucial to meeting expected timelines in automated guided vehicle operations. The main methodological contribution of this work lies in the integral process used to define an effective fitness function that guides the optimization method in the search for optimal solutions. This cost function plays a critical role in the effectiveness of the trajectory tracking algorithm by quantifying path quality and allowing comparisons between solutions. The design of this fitness function poses challenges including accuracy, suitability, minimization of path length, and avoiding or reducing collisions. To achieve the optimization objectives and address some issues such as sensitivity to parameter scaling and the risk of premature convergence, different approaches can be used. This work proposes to incorporate constraints into the fitness function, adjust the optimization parameters to reflect the conditions of the problem, and design a fitness function based on prior knowledge and an accurate representation of the goals. The three relevant contributions for the planning and optimization of routes of automated guided vehicle in industrial environments are the following. Firstly, the development of a mathematical model of trajectories based on Frenet curves that considers the static occupancy map of the environment. Second, an optimization strategy to generate optimal safe paths. Finally, a fitness function that guides the optimization method towards optimal solutions considering the sensitivity to scaling and resolution of the parameters. This study presents an exhaustive analysis of the different fitness functions obtained, each one evaluated based on key metrics such as the length of the trajectory, the average and minimum distance to the occupancy map, and the number of collisions along the path. The results show that the obtained cost function successfully avoids collisions with the environment in all scenarios and consistently remains the fitness function with the largest average distance to obstacles, at least 50% higher than other functions used in this study.

As spacecraft missions become more complex and ambitious, it becomes increasingly important to track the status and health of the spacecraft in real-time to ensure mission success. Anomaly detection is a crucial part of spacecraft telemetry analysis, allowing engineers to quickly identify unexpected or abnormal behaviour reflected on spacecraft data and take appropriate corrective action. Traditional statistical methods based on threshold setting are often inadequate for detecting anomalies in this context, requiring the development of more sophisticated techniques that can handle the high-dimensional, non-linear, and non-stationary nature of spacecraft telemetry data such as machine learning-based techniques. This article presents an approach for anomaly detection using machine-learning techniques for spacecraft telemetry. The identification of anomaly types present on two real telemetry datasets from NASA is performed to incorporate information of magnitude, frequency, and waveform from known anomalies into the feature extraction process. Then, a machine-learning-based model is trained with the obtained features and tested with unknown real data. The proposed method achieves 95.3% of precision and 100% of Recall, giving a F0.5 score of 96.2% in both datasets, outperforming the metrics obtained on the existing related works, demonstrating that the inclusion of known anomalies can improve the performance of the data-driven models. Finally, an explainability analysis is performed to understand why a particular data instance has been identified as anomalous, proving the effectiveness of the feature extraction process.

Abstract This paper presents an intelligent speed adaption system for vehicles on conventional roads. The fuzzy logic based expert system outputs a recommended speed to ensure both safety and passenger comfort. This intelligent system includes geometrical features of the road, as well as subjective perceptions of the drivers. It has been developed and checked with real data that were measured with an instrumental system incorporated in a vehicle, on several two‐lane roads located in the Madrid Region, Spain. Along with the road geometrical characteristics, other input variables to the system are external factors, such as weather conditions, distance to the preceding vehicle, tire pressure, and other subjective criteria, such as the desired comfort level, selected by the driver. The expert system output is the most suitable speed for the specific road type, considering real factors that may modify the category of the road and thus, the appropriate speed. This information could be added to the adaptive cruise control of the vehicle. The recommended speed can be a very useful input for both, drivers and the autonomous vehicles, to improve safety on the road system.

Abstract Fuel management is an important issue in aviation for safety and efficiency reasons. This work develops an alternative solution based on an artificial intelligent technique to the fuel distribution management problem in commercial aircrafts. The fuel flow control amongst tanks during the flight is addressed. A fuzzy management system is implemented that decides the best fuel distribution based on safety criteria (keeping engines fed) and dynamic stability (placing the centre of gravity at the appropriate position), amongst other specifications. Expert knowledge is used to define the rules of the fuel fuzzy control, taking into account that the dynamics of the system changes, whilst fuel is being consumed. It has been simulated on a real long‐range‐type commercial aircraft with satisfactory results regarding stability, even in the case of internal malfunction (in pipes, pumps, or valves), and with external disturbances (engine failure). The knowledge‐based fuzzy control is able to maintain the centre of gravity position within the stability and manoeuvrability margins along the flight. Besides, the intelligent control strategy minimizes the action of the actuators, providing some advantages to the control solution.

Abstract With the rapid growth of logistics transportation in the framework of Industry 4.0, automated guided vehicle (AGV) technologies have developed speedily. These systems present two coupled control problems: the control of the longitudinal velocity, essential to ensure the application requirements such as throughput and tag time, and the trajectory tracking control, necessary to ensure the proper accuracy in loading and unloading manoeuvres. When the paths are very short or have abrupt changes, the kinematic constraints play a restrictive role, and the tracking control becomes more challenging. In this case, advanced control strategies such as those based on intelligent techniques, including machine learning (ML) can be useful. Hence, in this work, we present an intelligent hybrid control scheme that combines reinforcement learning‐based control (RLC) with conventional PI regulators to face both control problems simultaneously. On the one hand, PIs are used to control the speed of each wheel. On the other hand, the input reference of these regulators is calculated by the RLC in order to reduce the guiding error of the path tracking and to maintain the longitudinal speed. The latter is compared with a PID path following controller. The PID regulators have been tuned by genetic algorithms. The RLC allows the vehicle to learn how to improve the trajectory tracking in an adaptive way and thus, the AGV can face disturbances or unknown physical system parameters that may change due to friction and degradation of AGV mechanical components. Extensive simulation experiments of the proposed intelligent control strategy on a hybrid tricycle and differential AGV model, that considers the kinematics and the dynamics of the vehicle, prove the efficiency of the approach when following different demanding trajectories. The performance of the RL tracking controller in comparison with the optimized PID gives errors around 70% smaller, and the average maximum error is also 48% lower.

Under the federated learning paradigm, the agents learn in parallel and combine their knowledge to build a global knowledge model. This new machine learning strategy increases privacy and reduces communication costs, some benefits that can be very useful for industry applications deployed in the edge. Automatic Guided Vehicles (AGVs) can take advantage of this approach since they can be considered intelligent agents, operate in fleets, and are normally managed by a central system that can run in the edge and handles the knowledge of each of them to obtain a global emerging behavioral model. Furthermore, this idea can be combined with the concept of reinforcement learning (RL). This way, the AGVs can interact with the system to learn according to the policy implemented by the RL algorithm in order to follow specified routes, and send their findings to the main system. The centralized system collects this information in a group policy to turn it over to the AGVs. In this work, a novel Federated Discrete Reinforcement Learning (FDRL) approach is implemented to control the trajectories of a fleet of AGVs. Each industrial AGV runs the modules that correspond to an RL system: a state estimator, a rewards calculator, an action selector, and a policy update algorithm. AGVs share their policy variation with the federated server, which combines them into a group policy with a learning aggregation function. To validate the proposal, simulation results of the FDRL control for five hybrid tricycle-differential AGVs and four different trajectories (ellipse, lemniscate, octagon, and a closed 16-polyline) have been obtained and compared with a Proportional Integral Derivative (PID) controller optimized with genetic algorithms. The intelligent control approach shows an average improvement of 78% in mean absolute error, 75% in root mean square error, and 73% in terms of standard deviation. It has been shown that this approach also accelerates the learning up to a 50 % depending on the trajectory, with an average of 36% speed up while allowing precise tracking. The suggested federated-learning based technique outperforms an optimized fuzzy logic controller (FLC) for all of the measured trajectories as well. In addition, different learning aggregation functions have been proposed and evaluated. The influence of the number of vehicles (from 2 to 10) on the path following performance and on network transmission has been analyzed too.

This work proposes a novel methodology to help in decision making in the cryptocurrency market. Two investment strategies have been designed for Ethereum (ETH), based on predictions of the price and trend of this cryptocurrency using real data. The two Ethereum cryptocurrency prediction systems rely solely on past values of other contextual stock indices, market indicators and online trends, and ignore any technical indicators of price evolution. Real data from cryptocurrency market has been collected and processed with different feature selection methods. Applying a regression approach with Gated Recurrent Unit (GRU) and Long Short-Term Memory (LSTM) networks, prediction models for the ETH price for 1, 7 and 15 days are obtained and compared. Also, support vector machine (SVM) is applied to predict the ETH price trend by applying a classification approach. In both approaches, sentiment analysis has been included to check its effect on the prediction results. The reliability of these prediction models in the current market has been evaluated by designing two original knowledge-based investment strategies. They are tested over two different time periods with real cryptocurrency market data. The results show that it is possible to generate up to 5.16 profit factor with few operations using these models. Furthermore, adding sentiment analysis has shown to have little influence. In this way, we contribute to the advancement of our knowledge of this volatile and still young cryptocurrency market, and specifically of the evolution of Ethereum and the factors that can influence its behavior.

Abstract In order to maintain adequate levels of cleanliness and sanitation in public facilities, prevent the buildup of viruses and other harmful pathogens, and ensure health and safety, health and labor authorities have repeatedly warned of the need to adhere to proper disinfection protocols in the workplace. This is particularly important in public places where food is handled, where there are more vulnerable people, including hospitals and health care centers, or where there is a large concentration of people. One promising approach is the combination of ultraviolet‐C (UV‐C) light and mobile robots to automate disinfection processes. Being this technology effective for disinfection, an excessive dose of UV can damage the materials, limiting its applicability. Therefore, a major challenge for automatic disinfection is to find a route that covers the entire surface, ensures cleanliness, and provides the correct radiation dose while preventing environmental materials from being damaged. To achieve this, in this paper a novel intelligent control approach is proposed. A bio‐inspired Glasius neural network with a motion planner, an UV estimation module, a speed regulator, and pure pursuit controller are combined into one intelligent system. The motion planner proposes a sequence of movements to go through the space in the most efficient way possible, avoiding obstacles of the environment. The speed controller adjusts the dose of UV‐C radiation and the pure pursuit regulator ensures the following of the path. This approach has been tested in various simulation scenarios of increasing complexity and in four different areas of dosing requirements. In simulation, a 44% reduction of the maximum dose is achieved, 17% less distance travelled by the robot and, what is more important, 229% more locations with the appropriate dose.

Wind turbine (WT) pitch control is a challenging issue due to the non-linearities of the wind device and its complex dynamics, the coupling of the variables and the uncertainty of the environment. Reinforcement learning (RL) based control arises as a promising technique to address these problems. However, its applicability is still limited due to the slowness of the learning process. To help alleviate this drawback, in this work we present a hybrid RL-based control that combines a RL-based controller with a proportional–integral–derivative (PID) regulator, and a learning observer. The PID is beneficial during the first training episodes as the RL based control does not have any experience to learn from. The learning observer oversees the learning process by adjusting the exploration rate and the exploration window in order to reduce the oscillations during the training and improve convergence. Simulation experiments on a small real WT show how the learning significantly improves with this control architecture, speeding up the learning convergence up to 37%, and increasing the efficiency of the intelligent control strategy. The best hybrid controller reduces the error of the output power by around 41% regarding a PID regulator. Moreover, the proposed intelligent hybrid control configuration has proved more efficient than a fuzzy controller and a neuro-control strategy.

Abstract This paper first presents a fuzzy expert system to identify and classify conventional two‐lane roads based on geometric characteristics. Both fuzzy and neuro‐fuzzy techniques have been used. Fuzzy logic has proved suitable to address this problem, since in this case, there is a variability of input information, and classical rules are not suitable to be used due to the uncertainty introduced by some combinations of the variables. Each road's geometric features were measured by sensors in an equipped vehicle, and are subsequently used to classify the roads according to their real condition. The conventional two‐lane roads used for this research are located in the Madrid Region, in Spain. This intelligent system may be used to update the road database regarding the assigned type to each conventional road, according to their present features and state. Also, a risk identification system has been developed to assess whether a vehicle is driving on a two‐lane road with an inappropriate speed, combining variables such as the former identification model, vehicle type, road longitudinal gradient, the angle covered by each horizontal curve, and the existence or not of an additional traffic lane. A fuzzy risk index is proposed for this approach. This fuzzy model may be useful to detect road sections where safety must be enhanced by revising the speed limit, since less safe situations may arise from travelling at unappropriated speeds.

Abstract This work focuses on the control of the pitch angle of wind turbines. This is not an easy task due to the nonlinearity, the complex dynamics, and the coupling between the variables of these renewable energy systems. This control is even harder for floating offshore wind turbines, as they are subjected to extreme weather conditions and the disturbances of the waves. To solve it, we propose a hybrid system that combines fuzzy logic and deep learning. Deep learning techniques are used to estimate the current wind and to forecast the future wind. Estimation and forecasting are combined to obtain the effective wind which feeds the fuzzy controller. Simulation results show how including the effective wind improves the performance of the intelligent controller for different disturbances. For low and medium wind speeds, an improvement of 21% is obtained respect to the PID controller, and 7% respect to the standard fuzzy controller. In addition, an intensive analysis has been carried out on the influence of the deep learning configuration parameters in the training of the hybrid control system. It is shown how increasing the number of hidden units improves the training. However, increasing the number of cells while keeping the total number of hidden units decelerates the training.

Abstract This work aims to establish a general and optimized procedure for the initial refuelling of commercial airplanes, as this loading process is strongly related to safety and energy saving issues. The on‐ground refuelling is addressed as an optimization problem whose cost function involves expert knowledge about constraints and factors that influence the aircraft stability and performance. Several heterogeneous criteria (fuelling time, structural load, flow transfers, etc.) have been considered and weighted accordance to its importance in terms of stability. This allows us to adapt the strategy to any type and planned trip of the airplane. The priority is the positioning of the centre of mass of the civil aircraft within safety and manoeuvrability margins, and near the optimal position. Evolutive algorithms are applied, keeping feasible solutions by modifying genetic operators. As a case of study, the initial refuelling of a long range type commercial aircraft, the Airbus A330‐200, is analysed. Simulation results have proved this methodology to be efficient and optimal. Even more, this heuristic and general approach improves the traditional solution that follows a set of pre‐defined rules that are specific for each type of aircraft.

In this study a new internal clustering validation index is proposed. It is based on a measure of the uniformity of the data in clusters. It uses the local density of each cluster, in particular, the normalized variability of the density within the clusters to find the ideal partition. The new validity measure allows it to capture the spatial pattern of the data and obtain the right number of clusters in an automatic way. This new approach, unlike the traditional one that usually identifies well-separated compact clouds, works with arbitrary-shape clusters that may be contiguous or even overlapped. The proposed clustering measure has been evaluated on nine artificial data sets, with different cluster distributions and an increasing number of classes, on three highly nonlinear data sets, and on 17 real data sets. It has been compared with nine well-known clustering validation indices with very satisfactory results. This proves that including density in the definition of clustering validation indices may be useful to identify the right partition of arbitrary-shape and different-size clusters.

This work analyses the use of parallel processing techniques in synthetic aperture ultrasonic imaging applications. In particular, the Total Focussing Method, which is a O(N2x P) problem, is studied. This work presents different parallelization strategies for multicore CPU and GPU architectures. The parallelization processes on both platforms are discussed and optimized in order to achieve real-time performance.

In this work, we are proposing a trading system where fuzzy logic is applied not only for defining the trading rules, but also for managing the capital to invest. In fact, two fuzzy decision support systems are developed. The first one uses fuzzy logic to design the trading rules and to apply the stock market technical indicators. The second one enhances this fuzzy trading system adding a fuzzy strategy to manage the capital to trade. Additionally, a new technical market indicator that produces short and long entry signals is introduced. It is based on the moving average convergence divergence indicator. Its parameters have been optimized by genetic algorithms. The proposals are compared to a classical nonfuzzy version of the proposed trading systems and to the buy-and-hold strategy. Results favor our fuzzy trading system in the two markets considered, NASDAQ100 and EUROSTOXX. Conclusions suggest that the use of fuzzy logic for capital management is promising and deserves further exploration.

The aim of this article is to analyse the results obtained when using a genetic algorithm (GA) to optimise the interplanetary trajectory of a spacecraft. The desired trajectory should visit Saturn, after performing a gravitational assistance or fly-by in planet Mars. The GA tunes a set of variables, in order to achieve the mission purpose while satisfying the constraints and minimizing the delta-V of the mission. Due to the complexity of the implemented models and the lack of analytical solutions, an alternative non-traditional algorithm provided by soft-computing techniques such as GA is necessary to achieve an optimum solution. The positions of planets as provided by Jet Propulsion Laboratory have been considered. A variable mutation rate has been implemented that broadens the search area whenever the population becomes uniform. The results are very useful from the point of view of mission analysis and therefore can be used as an initial guess for further optimizations. They can also be the first step for a more refined analysis and time-consuming simulations based on more complex models of orbital perturbations.

The control of nonlinear systems has been putting especial attention in the use of Artificial Intelligent techniques, where fuzzy logic presents one of the best alternatives due to the exploit of human knowledge. However, several fuzzy logic real-world applications use manual tuning (human expertise) to adjust control systems. On the other hand, in the Intelligent Transport Systems (ITS) field, the longitudinal control (throttle and brake management) is an important topic because external perturbations can generate uncomfortable accelerations as well as unnecessary fuel consumption. In this work, we utilize a neuro-fuzzy system to use human driving knowledge to tune and adjust the input-output parameters of a fuzzy if-then system. The neuro-fuzzy system considered in this work is ANFIS (Adaptive-Network-based Fuzzy Inference System). Results show several improvements in the control system adjusted by neuro-fuzzy techniques in comparison to the previous manual tuned controller, mainly in comfort and efficient use of actuators.

In this paper a controller based in the eigenvalues assignment technique is designed. The system to be stabilized is an unmanned helicopter. This eigenstructure based controller is designed considering all the measurable states. A LQ controller has also been applied to the same system in order to compare the responses regarding the robustness. The results are encouraging and prove that the eigenstructure assignment technique is useful for these nonlinear systems.

This simulation study has been developed to assess a set of realistic proposals of a new distribution centre to a pharmaceutical company, with a focus on the sorting operation. Different sorting strategies, establishing dynamic assignment of the lanes to destinations and lanes to operators are proposed. Also different packing procedures have been presented according to the operator way of working. New qualitative criterions have been defined to evaluate the strategies. The results show some interesting non-linear behaviour in the sorting operation, and give the number of required operators, length of the lanes, and working procedures to be used in this application. The simulation experiments show that the improvement in the overall productivity by choosing a specific sorting strategy can be significant. The study has helped the management of the involved company to make a decision about the supplier and actually, the suggested proposals are being implemented in practice.