High-Speed rail (HSR) infrastructures are developing all over the world. This represents a sustainable travel option according to sustainability goals. Recent works have proposed adopting the LARG (Lean, Agile, Resilient, Green) approach to realise new high-speed rail (HSR) lines. This is a paradigm derived from industry and logistics. The adoption of the LARG paradigm in transport systems and railway infrastructures has different implications. The main objective of this paper is to investigate the main implications of the Lean component for realising a new HSR line. The Lean paradigm focuses primarily on cost reduction by eliminating waste and activities that do not generate added value. In the context of High-Speed Rail systems, this approach directly influences both infrastructure design criteria and service planning. It requires comparing the requirements defined during the design phase with the actual, measurable outcomes observed during operation (ex post). This study focuses on the LEAN component applied to a case study relative to the extension of the Italian high-speed rail network along the southern corridor. The case study raises significant questions about the effectiveness of national infrastructure investments, emphasising the need for a broader discussion about the technical criteria used to develop existing high-speed rail lines. This work is of interest to the scientific community, as well as to public and private decision-makers involved in transportation planning connected to high-speed rail infrastructure.
Task scheduling in heterogeneous edge environments involves conflicting objectives related to completion time and energy consumption. Classical low-overhead heuristics usually optimize one dominant criterion, whereas population-based multi-objective methods provide richer trade-off analysis at a higher decision cost. This paper studies a batch scheduler, called avg_frontier, intended for repeated scheduling in such environments. The proposed method is organized as a two-stage procedure. First, it constructs an aggregated surrogate model of available devices using average time and energy coefficients over feasible tasks. The scheduler solves a discrete bi-objective allocation problem in the space of task counts, builds a surrogate Pareto frontier, and selects a compromise point according to the area_ratio policy controlled by α . Second, the selected allocation is restored into a real task-to-device assignment and re-evaluated by exact simulation. The algorithm is evaluated in SimGrid on two heterogeneous testbeds: a compact three-node profile with 60 tasks and a generated 100-node profile with 6000 tasks. The comparison includes classical heuristics and Pareto-based evolutionary methods. The results show that avg_frontier provides a controllable transition between time-oriented and energy-oriented regimes while keeping the scheduler overhead substantially lower than population-based Pareto baselines. The method is not intended to dominate all alternatives in every region of the trade-off, but to provide a computationally cheap mechanism for navigating this trade-off under repeated scheduling.
This paper proposes a new format for a digital Birth Certificate. It contains a color photograph of the child's face and facial biometric characteristics, documentary information (child's full name, year of birth, etc.), information about the educational institution, and the full names of the mother and father. The developed software enables the creation of a digital Birth Certificate in the form of a BIO QR code that stores all the necessary documentary information about the child, which can be read by standard barcode scanners and software. The BIO QR code is embedded in the LSB layers of the color image of the child's face, allowing for the concealment of information, as the resulting image with embedded data is indistinguishable from the original. The proposed digital Birth Certificate is a GIF file consisting of a set of multimedia files (graphic, biometric, and documentary) that compactly store information about the child. The Digital Birth Certificate is a document that can be updated at various stages of a child's development and allows for the adjustment of documentary information and the characterization of their facial biometrics.
The ongoing debate on proximity-based planning has produced a wide range of methods for assessing access to everyday services. Yet, most of these approaches remain centred on what can be reached within a given time threshold, while giving less attention to the spatial structure that makes some streets more strategically relevant than others. This paper argues that, particularly in suburban areas, streets should be evaluated for their structural strategic value. Special attention should be given to corridors that connect weakly linked sub-systems, as they could support future proximity-oriented interventions. Building on recent research about weak ties in spatial systems and the Kemeny-based centrality measure (KBC), this paper proposes a simple diagnostic approach to bridgeness, developed within the framework of the DUT-funded EMC2 project. The aim is not to compare alternative bridgeness metrics, but to examine whether high-KBC segments in a suburban area already coincide with local functions, or instead reveal structurally important yet underused corridors. The approach is tested in Turano, a suburban area of Massa, Italy, located within the wider Versilia conurbation. Using a road-centreline representation and a verified inventory of amenities, the study investigates the relationship between bridging structure and the distribution of Points of Interest through the concepts of alignment and mismatch. The results offer an exploratory but operational framework for distinguishing between corridors that already act as local functional anchors and latent spatial assets whose potential is not yet reflected in the current distribution of services. The paper concludes that bridgeness can provide a useful additional layer for proximity-oriented planning, particularly when combined with reliable local data and with broader assessments of walkability, public-space quality, and functional needs.
The spread of malicious information on social networks is often explosive and difficult to control, especially in small-world structures and echo chambers. Traditional intervention strategies based on degree centrality are effective in scale-free networks; however, they often perform poorly in relatively homogeneous community networks, where bridge nodes may be more important than hubs. In this paper, we propose SPID (Shortest-Path Interdiction for Diffusion), a corridor-based mitigation framework that targets fast-spreading “diffusion corridors”, defined as shortest paths connecting high-risk source regions to target communities. The experimental implementation uses SPID++, a residual-updated version of SPID that adapts to rerouted paths after each intervention step. Experiments on the Watts–Strogatz network show that SPID++ reduces the total number of infected nodes by approximately 2.2 times compared with Degree Centrality. Budget sensitivity analysis further shows that SPID++ can protect up to 444 more nodes than Degree as the intervention budget increases. Runtime results indicate that batched shortest-path computation reduces the intervention-selection time of SPID++ while preserving the selected intervention set.
In contemporary urban planning, fragmented suburban contexts and limited public resources require operational tools able to identify where public spaces and facilities can be introduced with limited conflict, cost, and land consumption. This paper proposes a parcel-based computational decision-support framework for mapping urban spatial potential in X-minute city planning. Taking cadastral parcels as the elementary unit of transformation, the model evaluates land suitability for two distinct targets – open public spaces and built public facilities – through a multi-criteria structure addressing morphology, regulatory framework, ownership and accessibility issues, lot coverage and uses, environmental and landscape constraints. The framework is applied to the suburban area of Massa (Italy), a heterogeneous territory characterized by strong spatial fragmentation, low service provision, and high dependence on private mobility. These conditions make it a particularly relevant testing ground for proximity-based planning approaches, where the challenge lies in activating and interconnecting dispersed spatial resources. Results highlight differentiated spatial potentials for public spaces and facilities, confirming that proximity planning cannot rely on a single transformation logic. At the same time, private interests and ownership emerge as critical structural constraints, emphasizing the need to explicitly integrate property conditions into computational planning tools. The study also outlines how parcel-level suitability can support a network-based interpretation of urban systems, enabling the identification of strategic nodes and connections. This work provides a foundation for future research aimed at adapting the model to different policy scenarios and integrating advanced network analytics, supporting more flexible and context-sensitive implementations of proximity-based planning.
Chaotic dynamical systems are widely used as entropy sources for key generation used for cryptographic purposes. This introduces a vulnerability: if an adversary were able to identify which chaotic system produced a given keystream, they would gain an important advantage for cryptanalysis. In this paper, we propose a deep learning framework for chaotic system fingerprinting, formulated in terms of a time-series classification problem over short keystream sequences. The model is trained and evaluated on sequences generated by four representative chaotic systems (Lorenz, Rössler, Sprott, Hénon 3D), using strict train/test separation. A regularized Residual Network (ResNet) model achieves over 92
Satellite data on pollutants have too coarse a spatial resolution to be suitable for use in studying the exposure of cultural heritage. The integrated use of environmental and anthropogenic data with pollution data derived from Sentinel-5p using machine learning methods makes it possible to obtain high-resolution data that can be used to map urban pollution, which is useful for assessing the potential exposure of cultural heritage to atmospheric degradation.
Internet of Medical Things (IoMT) is redefining modern healthcare by enabling the continuous collection, transmission, and analysis of patient data through interconnected medical devices, wearables, and hospital systems. These advantages, however, extend and complicate the e-health attack surface, the effects of which are devastating as they impact human health and privacy. Reports indicate a rise in IoT-based cyberattacks targeting medical infrastructures and sensitive data regarding patients’ health; consequently, Intrusion Detection and Prevention Systems (IDS/IPS) have become essential components of secure IoMT architectures. IoMT networks generally have different characteristics from traditional networks and require ad hoc solutions. Most IDS/IPS models in the literature rely on deep learning architectures that, while achieving high accuracy, are unsuitable for real-time operation in resource-constrained IoMT environments. The resulting gap between recognition performance and solution deployability leaves IoMT networks lacking real-world applicable approaches. To address this gap, this paper introduces a lightweight hybrid IDS/IPS framework that combines decision tree models with a compact neural network. The neural network is enhanced through a supervised binning mechanism designed to handle non-continuous and heterogeneous IoMT data effectively. Experiments on three widely adopted IoMT datasets (CICIoMT 2024, IoMT TrafficData, and WUSTL EHMS 2020) show that the proposed model achieves accuracy and F1-weighted scores exceeding 0.99, while processing 30,000–100,000 samples per second. Compared to state-of-the-art deep models, our approach maintains comparable or superior detection capability with significantly lower computational cost. The model proved to be particularly robust and effective in attack detection (binary classification) and attack classification (multiclass classification) tasks.
The rapid growth of the Internet due to the COVID pandemic and the increasing need for video content are creating pressure on existing network infrastructure. Therefore, there is a need for efficient Quality of Service routing techniques. In this context, this paper presents a new algorithm for Traffic Flow Planning system in SDN based on Ant Colony Optimization (TFPACO). In this algorithm, unlike existing optimization techniques, mice and elephant flows are considered separately. Mice flows are optimized using a Dijkstra algorithm considering risk and delay while elephant flows use a multi-objective ant colony optimization model. As a result we get multiple solutions, and only nondominated solutions are stored in the archive. Among the solutions obtained in the archive, a set of solutions is chosen for proportional routing. In this context, Dijkstra’s algorithm, multi-objective Dijkstra’s algorithm, TSACO, and the Motlagh algorithm are used as reference algorithms. A discrete-time simulation environment with dynamic edge parameters is used to check the performance of the proposed algorithm. The simulation results show that the proposed algorithm entirely eliminates flow and demand blocking with less latency than existing algorithms.
This work comes from a simple but urgent question: how can we produce clean energy without consuming our most precious resources, namely soil, landscapes and the memory of places? Photovoltaics is one of the great hopes of the energy transition. But when it is installed on the ground, transforming countryside, hills and natural areas into fields of panels, we risk replacing one problem with another. The landscape changes, the soil is lost, and local communities perceive it as an imposition rather than a shared transformation. Our contribution aims to change this point of view: what if all this energy could be generated without consuming new space, but by enhancing what already exists? The roofs of houses, agricultural sheds and industrial buildings can become a new “energy surface”, capable of producing without taking away, thus minimizing the impact on the landscape and land consumption. The research proposes an integrated method of spatial analysis and remote sensing that helps to design scenarios in which the ecological transition respects the meaning of places and does not pit energy against landscape, but allows them to coexist; in fact photovoltaics can only be sustainable if we start from a new perspective, capable of accompanying the use of technology.
High-Speed Railway (HSR) systems have established worldwide for their technical characteristics, which wherever they are built have allowed greater economic, social and environmental sustainability of transport. 60,000 km of HSR in operation and 60,000 km under construction or design testify to the increasing importance of HSR systems. For the planning of new lines and for the monitoring of the current ones, it is necessary to have knowledge of modal choice that allow us to evaluate the goodness of the infrastructure investments and the validity of the services offered to users. To this end, it is essential to collect data on users’ choices in the current situation. The overall results obtained provide valuable insights for understanding current transport flows and provide a basis for analysing future trends resulting from the introduction of new HSR lines in southern Italy. The effectiveness of the HSR system depends on the consistency between technical planning and design choices and international standards. Verifying this consistency is made possible by an in-depth analysis of demand, carried out using mobility data. The preliminary results obtained represent a first step in this direction.
The increase of carbon footprint and the climate change scenario are central themes of the current technical and political debate. An important contribution is provided by the transport of passengers and goods. Over long national distances the problem becomes even more crucial. In a nutshell, the carbon footprint of High-Speed Rail (HSR) is on average, lower than other modes of transport, such as private cars or airplanes, for the operating phases. The reduction of carbon emissions with the use of HSR for passengers’ trips at national scale would make it possible to pursue the Sustainable Development Goals (SDGs) included in the 2030 Agenda, with specific reference to Goal 13: “Climate change”. It is necessary to carry out an analysis of the current scientific literature on this topic, considering the levels of environmental impacts in relation to the operative phases of the HSR, through an analysis and comparison of the values of the indicators. The purpose of the paper is to give a synthetic answer to the problem: how sustainable are the HSR systems in terms of carbon footprint? The results will be twofold. On the one hand, parametric values can be defined which can synthetically represent the environmental advantage provided by the implementation of the HSR for passengers’ trips; on the other hand, the fields of unknowledge in which it is useful to carry out research can be delimited. The reference result is given by applying the synthetic parametric approach to a case study: the relationship Rome-Milan (Italy). The link is presented in relation to the current airline connections and passengers flow, and considering several scenarios connected with the implementation of the HSR line. The conclusions allow decision-makers and planners to be supported in the design and construction of new HSR lines with a broader point of view, which considers environmental CO2 emissions.
Maritime ports represent the primary gateway between a country’s land-based transport system and the global marketplace. Their capacity and configuration in relation to freight flow demands constitute a central topic of research within the field. However, vulnerabilities within the terrestrial transport network expose these ports to disruptive factors of both internal and external origin. In the case of Romanian maritime ports, the bridges crossing the Danube ensure their connection to the national road network, but they may also, in the event of external interference, lead to the isolation of these ports and a consequent reduction in the volume of goods passing through port terminals. The application of a macro-level network simulation model allows for the assessment of the impact generated by the unavailability of any of these critical road network components. The results obtained are useful for decision-makers in developing transport network strategies that ensure redundancy, with beneficial effects that can be economically quantified.
Autism spectrum disorders (ASD) do not have objective neurobiological markers, and diagnosis is based solely on behavioral scales. Functional resting MRI allows you to study disorders of brain connectivity, but existing approaches face problems of high dimensionality of data, overfitting and low interpretability. The article proposes a methodology for identifying stable, interpretable biomarkers of functional connectivity, reliably distinguishing individuals with ASD and neurotypical controls. To reduce dimensionality and ensure interpretability, two-stage feature selection (t-test with FDR correction, SelectKBest) and logistic regression are used. The model achieved an accuracy of 71.8
This study evaluates the efficiency of natural tourism resource use in Uzbekistan from the perspective of sustainable tourism development. The central premise of the research is that the dynamics of tourist flows reflect not only the overall scale of tourism activity, but also, indirectly, the level of utilization and economic significance of natural tourism resources. For this purpose, the total number of tourists was analyzed using annual time-series data for the period 2010–2024. This indicator was constructed as the sum of domestic tourists and inbound foreign visitors. To identify the future dynamics of tourist flows, the AutoRegressive Integrated Moving Average (ARIMA) model was employed. The Dickey–Fuller test confirmed that the series became stationary after second differencing. Based on the analysis of autocorrelation and partial autocorrelation functions, as well as the Akaike and Bayesian information criteria, the ARIMA(1,2,1) specification was selected as the most appropriate model. The findings demonstrate the growing economic importance of tourism based on natural tourism resources in Uzbekistan. At the same time, the projected increase in tourist flows implies a higher intensity of resource use, which makes rational management, stronger environmental monitoring, and scientifically grounded regional tourism planning increasingly necessary. The study substantiates tourist-flow forecasting as an important analytical instrument for assessing the efficiency of natural tourism resource use and for shaping sustainable tourism policy.
The increasing volatility of liner shipping networks, characterized by frequent capacity reallocations and schedule disruptions, calls for robust and comparable container routing models. This paper presents a systematic comparison of two approaches: the Connection Scan Algorithm (CSA) and the RAPTOR algorithm. To enable a model-independent assessment, routes are projected onto a canonical representation based on sequences of ports and liner services. Using a set of scenarios to increase network complexity, routing outcomes, computational performance, and route similarity between the two approaches are evaluated. The results show that CSA and RAPTOR exhibit greater sensitivity to network dynamics, generating structurally different routes with more transshipments, particularly in large and volatile scenarios. The proposed framework provides a unified basis for comparing heterogeneous routing models and highlights trade-offs between computational efficiency, temporal realism, and robustness, offering guidance for both strategic planning and operational analysis in modern liner shipping networks.
This paper addresses the reconstruction of images from xenographic documents, which often exhibit multiple and varied types of distortions. To effectively resolve these issues, a multi-stage reconstruction process is required, with each stage targeting a specific distortion. A primary challenge in this process is the segmentation of pixels according to the printer’s color palette. This study proposes incorporating the segmentation phase at an early stage of reconstruction and introduces a technique to achieve this objective. The experimental results demonstrate promising outcomes for facilitating subsequent reconstruction stages.
This work addresses the parameter estimation of a fractional viscoelastic model comprising a springpot and a fractional dashpot arranged in series. Key features of the model, including the relaxation modulus, are formulated in the Laplace domain and subsequently evaluated in the time domain through numerical inversion using the Talbot method. A two-step optimisation strategy is adopted to identify the model parameters efficiently. In the first stage, frequency-domain data, namely the storage and loss moduli, are used to estimate an initial set of parameters. In the second stage, these preliminary estimates are refined by performing an optimisation that incorporates both frequency-domain data and time-domain data (relaxation modulus). Since the relaxation modulus is obtained through an inverse Laplace transform and does not admit a closed-form expression in the time domain, a sequential iterative procedure is employed during the second stage. In this approach, the relaxation modulus is first computed numerically using the Talbot method, after which the parameters are updated through optimisation. This process is repeated iteratively until a prescribed tolerance is satisfied. Validation against experimental data obtained for a low-density polyethylene demonstrates that the proposed methodology accurately reproduces both frequency- and time-domain responses, highlighting the effectiveness of fractional models in capturing complex viscoelastic behaviour.
This paper addresses the problem of image reconstruction in the presence of different types of noise, focusing on the analysis and comparison of several filtering techniques. In particular, Neural Network (NN) operators and sampling Kantorovich (SK) operators are investigated alongside a variety of classical and advanced filtering methods, including linear, nonlinear, and state-of-the-art approaches commonly adopted in the literature. The performance of the considered methods is evaluated through widely used quantitative metrics. The experimental analysis highlights the ability of NN and SK operators to provide accurate and stable reconstructions, ensuring the effectiveness and versatility of the proposed operators across different noise conditions.
