Urban warfare presents complex and dynamic challenges for modern military operations, demanding adaptive and responsive support systems. This study presents a Battle Management System (BMS) based on a Wireless Sensor Network (WSN) integrated with an Artificial Intelligence (AI)-powered planner to address key operational issues such as task coordination, resource allocation, mobility management, energy optimization, and communication resilience. Unlike prior works, our study introduces a tightly coupled integration where real-time sensor data dynamically influences AI-generated plans, allowing mission adaptation under battlefield constraints. Our study covers the practical implementation of this AI-enhanced BMS, a showcase of its capabilities through comprehensive urban warfare simulations, its real-world validation at a military facility demonstrating its effectiveness in an operational setting, and a performance benchmarking against state-of-the-art planners, showing superior results in mission success, risk reduction, and resource efficiency. Our findings confirm that AI-driven planning, in conjunction with WSNs, can generate operational plans that address the unique demands of urban combat scenarios with high precision and efficiency.

Background: In the field of Automated Planning, the existence of multiple agents in a certain temporal setting introduces the possibility of concurrency. This severely increases the complexity of planning in multi-agent temporal scenarios, as the possible states in the search space grow exponentially. Objectives: These types of domains are traditionally solved by making use of temporal reasoning techniques that do not directly address the “multi-agent nature” at their core. In contrast, we introduce MA-LAMA, a multi-agent temporal planner that makes use of multi-agent techniques to deal with the inherent complexity of multi-agent temporal scenarios. Methods: We propose a sequenced framework in which several multi-agent planning techniques are applied: automatic agent detection, task decomposition, cost-informed goal assignment, and agent interaction analysis; that are aimed to reduce the search complexity when solving temporal planning tasks. Results: Our results show that, in many cases, MA-LAMA outperforms other state-of-the-art temporal planners in plan cost optimization for well-known temporal domains. Conclusions: These results, along with the fact that MA-LAMA does not incorporate any temporal reasoning during search, suggest that several widely considered temporal domains are best suited to be solved with multi-agent planning techniques, rather than with temporal reasoning.

OBJECTIVE: This paper aims to develop a data-driven simulation framework for modeling patient flow in a hospital Emergency Department using interpretable methods throughout the entire process in the absence of system resource data. The goal is to improve understanding of system dynamics and support decision-making processes through transparent simulations, even when resource data are unavailable. METHODS: We developed a simulation framework using anonymized medical records from a Spanish hospital's Emergency Department. The model captures patient flow considering triage levels by identifying routes and measuring the transition times between each stage in them. We estimated these transitions using both parametric (theoretical) distributions and non-parametric Kernel Density Estimation (KDE). Patient admissions times are modeled by using probability distributions. We enhanced realism through an iterative refinement process guided by tolerance thresholds and quantitative metrics. This process refined the synthetic data to match the original distributions. RESULTS: Our approach produces highly realistic patient flow simulations with low tolerance values in the iterative method. The process gradually converges toward the original data. Distance and divergence metrics, together with statistical test results, indicate a high degree of similarity between the simulations and the real data, passing the Mann-Whitney U and Kolmogorov-Smirnov tests simultaneously in 100% of the generated samples when the tolerance threshold is low. CONCLUSION: The experimental results demonstrate that our simulation method effectively reproduces patient flow dynamics with a high level of realism and flexibility, even in the absence of information related to service resources. Its interpretable design and adjustable parameters enable safe data analysis and the exploration of alternative management strategies (e.g., modifying potential patient routes or restricting some transitions). These features position the methodology as a valuable tool for supporting informed decision-making and suggest its potential for use in other hospitals with suitable data, pending validation on external datasets.

Several multi-agent techniques are utilized to reduce the complexity of classical planning tasks, however, their applicability to temporal planning domains is a currently open line of study in the field of Automated Planning. In this paper, we present MA-LAMA, a factored, centralized, unthreated, satisfying, multi-agent temporal planner, that exploits the 'multi-agent nature' of temporal domains to perform plan optimization. In MA-LAMA, temporal tasks are translated to the constrained snap-actions paradigm, and an automatic agent decomposition, goal assignment, and required cooperation analysis are carried out to build independent search steps, called Search Phases. These Search Phases are then solved by consecutive agent local searches, using classical heuristics and temporal constraints. Experiments show that MA-LAMA is able to solve a wide range of classical and temporal multi-agent domains, performing significantly better in plan quality than other state-of-the-art temporal planners.

Emergency Department overcrowding is a well-known problem. The consequences are long waiting times for patients, reduced service quality, and the potential for increased mortality rates. Moreover, this problem is escalating with the aging of the population and the chronification of diseases. Being aware of such concerns, Emergency Departments started to conduct patient accountability years ago, and now they are ready to take advantage of the gathered data to improve the quality of service to citizens. However, to achieve this goal, they need the technology that enables the generation of predictive models from such data to improve the planning of their resources. In this paper, we face the task of forecasting patient admissions by using Deep Neural Networks, particularly a modern Attention-based model. As well, historical records do not provide a full picture of how patients attend the emergency service. Relying on the possibilities given by the attention mechanism, we also propose the use of exogenous information such as calendar data, weather, air quality, allergens, and information extracted from the web via Google Trends to enhance admissions prediction accuracy. In this regard, we have tested different configurations of the exogenous variables to isolate which ones provide relevant information that improves the model. In our experiments, we have seen that the calendar, weather, and air quality provide the most valuable information, meanwhile, allergens and Google Trends data appear to be hindering the models’ performance.

Military tactical scenarios have been shifting to more often consider combat situations in urban environments. Threats in these environments are generally more dynamic in nature, imposing new requirements on sensors and communications systems that support military operations. Wireless sensor networks (WSNs) with a large number of small and mobile computing nodes became the typical solution. However, WSNs demand additional complexity to dynamically manage their tasks, resource allocation, mobility, power consumption, and communication. This paper illustrates the integration of AI techniques into a Battle Management System (BMS) to support military operations in urban environments. The BMS is enhanced with an AI-based planner able to plan tasks, allocate resources, and monitor the WSN operation. The planner takes into consideration energy harvesting capabilities, secure data transfer, and authorization procedures. It generates plans using the information received from the sensors. In case new situations emerge, based on data fusion information, it automatically replans to adapt to the uncertainty in the environment. Finally, it takes into account the coverage between the different components to optimize the communications and better support WSN’s operator(s) and their activities.

Abstract Hospitals’ Emergency Departments (ED) have a great relevance in the health of the population. Properly managing the ED department requires to optimise the service, while maintaining a high quality care. This trade-off implies to properly arrange the schedule for the personnel, so the service can duly attend all patients. In this regard, a key point is to know in advance how many patients will arrive to the service and the number that should be derived to hospitalisation. To provide such information, we present the results of applying different algorithms for forecasting ED admissions and hospitalisations for both seven days and four months ahead. To do this, we have employed the ED admissions and inpatients series from a Spanish civil and military hospital. The ED admissions have been aggregated on a daily basis and on the official workers’ shifts, while the hospitalisations series have been considered daily. Over that data we employ two algorithms types: time series (AR, H-W, SARIMA and Prophet) and feature matrix (LR, EN, XGBoost and GLM). In addition, we create all possible ensembles among the models in order to find the best forecasting method. The findings of our study demonstrate that the ensembles can be beneficial in obtaining the best possible model.

In this article, we present fundamental design flaws of CaptchaStar. We also present a full analysis using the BASECASS methodology that employs machine learning techniques. By means of this methodology, we find an attack that bypasses CaptchaStar with almost 100% accuracy.

In a smart environment like the smart grids, it is necessary to have knowledge models that allow solving emerging problems. However, large datasets are required to automatically create these models, which in the vast majority of cases are not available. Therefore, in these environments is essential to have a data generator for each context. In this paper, we propose a synthetic data generation system based on the variational autoencoder (VAE) technique and linked data paradigm, to create larger datasets from small datasets acting as samples. Specifically, a Linked Data-based dataset extractor is proposed, which allows obtaining samples of data in a particular context. Then, a VAE is trained with these samples of data available in a specific context, to learn the latent distribution that characterizes the dataset, allowing the production of new records that are similar to those of the original dataset. Finally, several case studies in the field of energy management are carried out. In one of them, the process that follows our approach is described in detail; and then, another 3 more are considered to evaluate its ability to automate the data generation process for smart energy management. The results show how our synthetic data generation system can be used to obtain synthetic datasets in different energy contexts.

Currently, there is a high rate of generation of new information about the Energy Consumption of customers. It is important the traceability of its consumption pattern evolution to determine in real-time the services of a smart energy management system. This paper analyses the evolution of the Energy Consumption Pattern of customers using the Learning Algorithm for Multivariable Data Analysis (LAMDA). LAMDA is a fuzzy approach for supervised and unsupervised learning, based on the calculation of the Global Adequacy Degree (GAD) of one individual to a class/cluster, through the contributions of all its descriptors. LAMDA can create new classes/clusters after the training stage (online learning). If an individual does not have enough similarity to the preexisting classes/clusters, it is evaluated with respect to a threshold called the Non-Informative Class (NIC) to define if it is a new class/cluster. Particularly, the algorithm of the LAMDA family used in this paper is LAMDA-RD (Robust Distance). In the paper is analyzed the patterns of the initial grouping of the data, as well as, the patterns through their evolution (traceability). For the analysis of the patterns different metrics are considers: Calinski- Harabasz Index and Silhouette Score.

In this paper, we propose a supervised multilabel algorithm called Learning Algorithm for Multivariate Data Analysis for Multilabel Classification (LAMDA-ML). This algorithm is based on the algorithms of the LAMDA family, in particular, on the LAMDA-HAD (Higher Adequacy Grade) algorithm. Unlike previous algorithms in a multi-label context, LAMDA-ML is based on the Global Adequacy Degree (GAD) of an individual in multiple classes. In our proposal, we define a membership threshold (Gt), such that for all GAD values above this threshold, it implies that an individual will be assigned to the respective classes. For the evaluation of the performance of this proposal, a solar power generation dataset is used, with very encouraging results according to several metrics in the context of multiple labels.

Today, much of the interaction between clients and providers has moved to the Internet. Some tricksters have also learned to benefit from this new situation. New improved cons, tricks and deceptions can be found on-line. Many of these deceptions are only profitable if they are done at a large scale. In order to achieve these large numbers of interactions, these attacks require automation. CAPTCHAs/HIPs are a relatively new security mechanism against automated attacks. They try to detect when the other end of the interaction is a human or a computer program (a bot). However, CAPTCHA/HIP design is still in its initial conception as the stream of successful attacks highlight it. This paper focuses on the design of CAPTCHAs and if there is a way in which to assess a basic level of security for new CAPTCHA designs. To do so, we first review main attacks to different types of CAPTCHAs and then, we describe BASECASS, a methodology that can help in avoiding some of these design pitfalls. The application of the methodology is exemplified in three attacks to CAPTCHAs and how following the methodology designers could have avoided them.

Falls are the predominant cause of injury for older people. How to detect them is being in the last decade the focus of attention of many projects and researchers. This paper presents a simulator for recreating triaxial accelerometer measures of people falling in two circumstances: as a consequence of a loss of conscience or due to bumping into an obstacle. The objective of the simulator is to generate falling instances to train Machine Learning algorithms that can be incorporated into wearable devices. The developed simulator can generate triaxial accelerometer measures that exhibit similar patterns compared to falls recorded with real people and mannequins using a commercial device.

Human Interactive Proofs (HIPs 1 or CAPTCHAs 2 ) have become a first-level security measure on the Internet to avoid automatic attacks or minimize their effects. All the most widespread, successful or interesting CAPTCHA designs put to scrutiny have been successfully broken. Many of these attacks have been side-channel attacks. New designs are proposed to tackle these security problems while improving the human interface. FunCAPTCHA is the first commercial implementation of a gender classification CAPTCHA, with reported improvements in conversion rates. This article finds weaknesses in the security of FunCAPTCHA and uses simple machine learning (ML) analysis to test them. It shows a side-channel attack that leverages these flaws and successfully solves FunCAPTCHA on 90% of occasions without using meaningful image analysis. This simple yet effective security analysis can be applied with minor modifications to other HIPs proposals, allowing to check whether they leak enough information that would in turn allow for simple side-channel attacks.

Due to the complexity of theoretical approaches in evolutionary computation (EC), research has being largely performed on experimental basis. One popular measure used by the EC community is the success rate (SR), which is used alone or as part of more complex measures such as Koza’s computational effort in genetic programming. A common practice in EC is to report just a punctual estimation of the SR, without additional information about its associated uncertainty. We aim to motivate EC researchers to adopt more rigorous practices when working with SRs. In particular, we introduce the importance of correctly reporting this measure and highlight its binomial nature. Unfortunately, this fact is usually overlooked in the literature. Considering the binomiality of the SR opens the whole corpus of binomial statistics to EA research and practice. In particular, we focus on studying several methods to compute SR confidence intervals, the factors that determine their quality in terms of coverage probability and interval length. Due to its practical interest, we also briefly discuss the number of required runs to build confidence intervals with a certain quality, providing a sound method to set the number of runs, one of the most important experimental settings in EC. Evidence suggests that Wilson is, on average, a reliable and simple method to bound an estimation of SR with confidence intervals, while the standard method, which is quite popular because of its conceptual simplicity, should be avoided in any case. However, other methods can also be of interest under certain circumstances. We encourage to report the number of trials and successes, as well as the interval, to ease further comparability of the results.

Summary Human interactive proofs (HIPs) are a basic security measure on the Internet to avoid automatic attacks. There is an ongoing effort to find a HIP that is secure enough yet easy for humans. Recently, a new HIP has been designed aiming at higher security: the Civil Rights Completely Automated Public Turing test to tell Computers and Humans Apart (CAPTCHA). It employs the empathy capacity of humans to further strengthen Securimage, a well‐known text CAPTCHA. In this paper, we analyze it from a security perspective, finding fundamental design flaws. Using several well‐known machine learning (ML) algorithms, we analyze to what extent these flaws affect its security. We discover that thanks to them, we can create a successful side‐channel attack. This attack is able to correctly solve the HIP on 20.7%of occasions, much more than enough to consider it broken. Thus, we show that there is no need to solve the problem of optical character recognition nor empathy analysis for computers to break this particular HIP. ML can be successfully used to break a HIP that uses both with a side‐channel attack. This security analysis can be applied to other HIPs. It will allow to test whether they areleakingtoo much information by unexpected ways, given non‐evident design flaws. Copyright © 2015 John Wiley & Sons, Ltd.

Path-planning is a well-known studied problem in Artificial Intelligence. Given two points in a map, path-planning algorithms search for a path that joins those two points, avoiding obstacles. It is a challenging problem with important practical applications in a wide range of applications: autonomous mobile robotics, logistics or video games, just to mention some of them. Given its importance, it has attracted much research, resulting in a large number of algorithms, some classical, such as A*, other more specialized, such as swarms. However, despite all the literature dedicated to this problem, the statistics used to analyze experimental results in most cases are naïve. In this paper, we position in favor of the need of incorporating stronger statistical methods in path-planning empirical research and promote a debate in the research community. To this end, we analyze some 2D-grid classical path-planning algorithms in discrete domains (i.e. A* and A* with post-processing) and more recent algorithms in continuous domains (i.e. Theta* and S-Theta*). Given the differences of these algorithms, we study them under different criteria: Run-time, number of heading changes, number of expanded vertices and path-length.

The problem of Mission Planning for a large number of Unmanned Air Vehicles (UAV) can be formulated as a Temporal Constraint Satisfaction Problem (TCSP). It consists on a set of locations that should visit in different time windows, and the actions that the vehicle can perform based on its features such as the payload, speed or fuel capacity. In this paper, a temporal constraint model is implemented and tested by performing Backtracking search in several missions where its complexity has been incrementally modified. The experimental phase consists on two different phases. On the one hand, several mission simulations containing (n) UAVs using different sensors and characteristics located in different waypoints, and (m) requested tasks varying mission priorities have been carried out. On the other hand, the second experimental phase uses a backtracking algorithm to look through the whole solutions space to measure the scalability of the problem. This scalability has been measured as a relation between the number of tasks to be performed in the mission and the number of UAVs needed to perform it.

Abstract Radio frequency identification (RFID) is a powerful technology that enables wireless information storage and control in an economical way. These properties have generated a wide range of applications in different areas. Due to economic and technological constrains, RFID devices are seriously limited, having small or even tiny computational capabilities. This issue is particularly challenging from the security point of view. Security protocols in RFID environments have to deal with strong computational limitations, and classical protocols cannot be used in this context. There have been several attempts to overcome these limitations in the form of new lightweight security protocols designed to be used in very constrained (sometimes called ultra‐lightweight) RFID environments. One of these proposals is the David–Prasad ultra‐lightweight authentication protocol. This protocol was successfully attacked using a cryptanalysis technique named Tango attack. The capacity of the attack depends on a set of boolean approximations. In this paper, we present an enhanced version of the Tango attack, named Genetic Tango attack, that uses Genetic Programming to design those approximations, easing the generation of automatic cryptanalysis and improving its power compared to a manually designed attack. Experimental results are given to illustrate the effectiveness of this new attack.

Abstract Development of real planning and scheduling applications often requires the ability to handle uncertainty. Often this uncertainty is represented using probability information on the initial conditions and on the outcomes of actions. In this paper, we describe a novel probabilistic plan graph heuristic that computes information about the interaction between actions and between propositions. This information is used to find better relaxed plans and to compute their probability of success. This information guides a forward state space search for high probability, non‐branching seed plans. These plans are then used in a planning and scheduling system that handles unexpected outcomes by runtime replanning. We briefly describe the heuristic, the search process, and the results on different domains from recent international planning competitions. We discuss the results of this study and some of the issues involved in advancing this work further.

Some commonly used performance measures in Genetic Programming are those defined by John Koza in his first book. These measures, mainly computational effort and number of individuals to be processed, estimate the performance of the algorithm as well as the difficulty of a problem. Although Koza's performance measures have been widely used in the literature, their behaviour is not well known. In this paper we study the accuracy of these measures and advance in the understanding of the factors that influence them. In order to achieve this goal, we report an empirical study that attempts to systematically measure the effects of two variability sources in the estimation of the number of individuals to be processed and the computational effort. The results obtained in those experiments suggests that these measures, in common experimental setups, and under certain circumstances, might have a high relative error.

The new advances in information communications technology (ICT) allow expanding the electronic communication networks into an objective environment. These techniques can be used to help users who have to go to a building with several floors, corridors, and departments to get its right location and orientation. The problem gets more complicated when there are a high number of visitors with temporal constraints. Typically the lack of information and signs to guide the user complicates this task. A typical example is a hospital, where patients have scheduled doctor's appointments and in some cases movement problems (handicaps). A possibility for solving this problem is to equip the building with an intelligent system for user detection and orientation. In this article we present a solution that uses a detection and a location system based on wireless technology and artificial intelligence techniques to plan and inform about the paths the user can follow.

Success Rate (SR) is a statistic straightforward to use and interpret, however a number of non-trivial statistical issues arises when it is examinated in detail. We address some of those issues, providing evidence that suggests that SR follows a binomial density function, therefore its statistical properties are independent of the flavour of the Evolutionary Algorithm (EA) and its domain. It is fully described by the SR and the number of runs. Moreover, the binomial distribution is a well known statistical distribution with a large corpus of tools available that can be used in the context of EC research. One of those tools, confidence intervals (CIs), is studied.

Experience developing and deploying model-based diagnosis (MBD) and recovery and other model-based technologies on a variety of testbeds and flight experiments led us to explore why our expectations about the impact of MBD on spacecraft operations have not been matched by effective benefits in the field. By MBD, we mean the problem of observing a mechanical, software, or other system and determining what failures its internal components have suffered using a generic inference algorithm and a model of the system's components and interconnections. These techniques are very attractive, suggesting a vision of machines that repair themselves, reduced costs for all kinds of endeavors, spacecraft that continue their missions even when failing, and so on. This promise inspired a broad range of activities, including our involvement over several years in flying the Livingstone and L2 onboard MBD and recovery systems as experiments on Deep Space 1 and Earth Observer 1 spacecraft. Yet, in the end, no spacecraft project adopted the technology in operations nor flew additional flight experiments. To our knowledge, no spacecraft project has adopted any other MBD technology in operations. In this paper, we present a cost/benefit analysis for MBD using expectations and experiences with Livingstone as an example. We provide an overview of common techniques for making spacecraft robust, citing fault protection schemes from recent missions. We lay out the cost, benefit, and risk advantages associated with onboard MBD and use the examples to probe each expected advantage in turn. We suggest a method for evaluating a mission that has already been flown and providing a rough estimate of the maximum value that a perfect onboard diagnosis and recovery system would have provided. By unpacking the events that must occur in order to provide value, we also identify the factors needed to compute theexpectedvalue that would be provided by a real diagnosis and recovery system. We then discuss the expected value we would estimate that such a system would have had for the Mars Exploration Rover mission. This has allowed us to identify the specific assumptions that made our expectations for MBD in this domain incorrect.

This article presents a new type of Frequently Asked Questions (FAQ) System, called DynJAQ (Dynamic Java Asked Questions) that has been designed with the purpose of making learning more appealing to beginner students of engineering disciplines and overcome the inconvenience of these systems. DynJAQ is able to generate dynamically several HTML guides that can be used to answer any possible question about a particular programming language (Java), although it can be easily extended to any other topic. DynJAQ integrates case-based knowledge into a graph-based representation that can be easily learned and managed. The combination of both case-based knowledge and graphs allows it to implement a flexible hierarchical structures (or learning graphs) that have been applied to implement a new kind of Frequently Asked Questions Systems. In these systems the output is dynamically built from the user query, using as basis structures the knowledge retrieved from a Case Base. The management of these cases allows enriching the knowledge base. © 2007 Wiley Periodicals, Inc. Int J Int Syst 22: 303–318, 2007.

Recently, the areas of planning and scheduling in artificial intelligence (AI) have witnessed a big push toward their integration in order to solve complex problems. These problems require both reasoning on which actions are to be performed as well as their precedence constraints (planning) and the reasoning with respect to temporal constraints (e.g., duration, precedence, and deadline); those actions should satisfy the resources they use (scheduling). This paper describes IPSS (integrated planning and scheduling system), a domain independent solver that integrates an AI planner that synthesizes courses of actions with constraint-based techniques that reason based upon time and resources. IPSS is able to manage not only simple precedence constraints, but also more complex temporal requirements (as the Allen primitives) and multicapacity resource usage/consumption. The solver is evaluated against a set of problems characterized by the use of multiple agents (or multiple resources) that have to perform tasks with some temporal restrictions in the order of the tasks or some constraints in the availability of the resources. Experiments show how the integrated reasoning approach improves plan parallelism and gains better makespans than some state-of-the-art planners where multiple agents are represented as additional fluents in the problem operators. It also shows that IPSS is suitable for solving real domains (i.e., workflow problems) because it is able to impose temporal windows on the goals or set a maximum makespan, features that most of the planners do not yet incorporate