Road surface abrasions significantly contribute to vehicle collisions and mechanical failures worldwide. Traditional machine learning-based methods for road damage detection typically rely heavily on extensive manual annotations, making them costly, labour-intensive, and inefficient. To address this challenge, this paper introduces a label-efficient self-supervised learning framework designed to facilitate efficient, scalable, and automated detection of road surface defects. Our approach integrates contrastive learning with a regularized redundancy reduction method, enabling the extraction of rich, discriminate features directly from unlabelled data. Contrastive learning separates positive and negative samples to learn robust feature representations, while a cross-correlation loss maximizes information content by minimizing redundancy. Regularization through variance and covariance loss terms ensures feature diversity and prevents informational collapse in the learned representations. Extensive evaluations in both in-domain and cross-domain scenarios demonstrate that our proposed method achieves superior performance compared to supervised techniques, even when trained with substantially fewer labelled samples. Thus, this work provides an effective, economical, and scalable solution to the critical challenges faced in automated road maintenance.

This study presents a comprehensive multi-objective transportation model aimed at optimizing complex vehicle routing problems, which are nondeterministic polynomial time NP-hard due to spatial, temporal, and capacity constraints. In this study, the multi-objective transportation model integrates decisionmaker preferences with hybrid optimization techniques, including the approximatecombinatorial method, ant colony optimization and evolutionary algorithms. it seeks to minimize transportation costs, time, and emissions while accounting for real-world constraints such as fleet composition, customer demand, and servicelevel agreements. The techniques like multi-criteria decision-making methods are employed to refine the solution set, balancing objectives like cost, time, environmental impact, and service level. The novel optimization model is applied to a fuel distribution case study involving 18 customers and a heterogeneous fleet, where it optimizes vehicle routes to meet delivery requirements efficiently. The multiobjective transportation framework generates multiple feasible solutions, which are further narrowed down using decision-making frameworks to ensure alignment with organizational goals and decision-maker preferences. The integration of quantitative optimization techniques with qualitative decision-making processes makes this model robust and scalable, offering a practical tool for enhancing operational efficiency in transportation systems. This approach effectively addresses real-world logistics challenges, demonstrating significant improvements in route efficiency, cost savings, and environmental sustainability.

This paper presents a novel tool for optimising residential parking allocation in urban environments using linear programming techniques. The tool addresses the growing challenge of parking space management in cities by quantifying parking utilisation and accessibility. It employs a unique application of the transport problem from Graph Theory to allocate parking supply to household demand while considering real-world constraints such as walking distances and infrastructure limitations. The methodology involves the pre-processing of supply, demand, and distance matrix data, followed by an optimization process that minimises total walking distance and penalises unmet demand. The tool’s effectiveness is demonstrated through an experiment in the Czech town of Slany, showcasing its ability to evaluate current parking situations and assess the impact of potential changes in parking supply. Key outputs include the percentage of satisfied demand, utilization rates of parking supply, and detailed allocation maps. This approach provides urban planners and policymakers with valuable insights for developing efficient and sustainable parking solutions, while also highlighting areas for further research in data preparation and model refinement.

The primary objective of the presented research is to enhance an existing data quality control application by integrating advanced anomaly detection mechanisms based on generalized additive models. This approach targets time- series traffic data, where traditional methods may fall short in identifying complex, non-linear patterns of anomalies. In collaboration with Simplity s.r.o., we are extending their current data quality assessment tool to incorporate generalized additive models, providing a more robust and dynamic solution for monitoring and ensuring the reliability of traffic datasets. The integration of these models aims to improve the accuracy of anomaly detection, leading to more effective data management in transport systems and contributing to higher standards of data quality in the field of traffic informatics.

This paper aims to extract and select the significant features of electroencephalogram (EEG) signals and classify the visual stimulation of distinct colors. In this work, a novel method for selecting distinct colors using EEG signals called affinity artificial immune and Daubechies wavelet time-based learning (AAIDWTL) is proposed. Initially, the EEG signals were collected in a controlled environment and an in-built band-pass filter was applied to remove the artifacts. The filtered signals were converted into frequency domain signals using least squarebased short-term Fourier transform. After that, by utilizing Daubechies wavelet statistical time-based feature extraction model the time domain features were extracted. Followed by, computationally efficient features were selected using an affinity artificial immune-based feature selection model. The selected features were classified using a polynomial kernel multiclass classification-based machine learning algorithm and achieved an accuracy of 97.5% when compared with other methods like linear discriminant analysis (LDA) which obtained only 92%. Furthermore, while utilizing the proposed method classification time was considerably less when compared to LDA. The experimental result shows that the proposed color stimulation of the EEG signals method achieved greater improvement in terms of both classification time and classification accuracy with a minimum false positive rate.

Early detection of cervical cancer can help patients obtain the best treatment through various means. In general, computer-aided diagnosis has a high impact on the accuracy, reliability, and convenience of cervical cancer. However, several limitations have been faced through the design process in detecting or classifying the cells, such as variation of image features and low-image resolution. Moreover, shape indifference is one of the limitations in terms of image processing scope. The metrics used to measure the size and shape of the cells have not been developed to distinguish the differences between the shape of the objects. This paper focused on the detection and segmentation of the nucleus cell region in Pap smear images based on Bradley local thresholding. The proposed method evolved several steps, such as color adjustment, k-means, and a Bradley modification algorithm. Based on image quality assessment (IQA), the numerical evaluation results indicate that the proposed approach has segmented a full area of the nucleus cell region significantly and efficiently compared to the original Bradley algorithm. We obtained F-measure (98.62%), sensitivity (99.13%), and accuracy (97.96%). It has also been proven that the proposed method can effectively address the issue of low contrast and black noise. Hence, the proposed method differs from the previous research in terms of color disproportion adjustment and the modification of Bradleys algorithm for Pap smear image convenience.

This article addresses the challenges operators face in decision-making during the operational management of tunnels and other transport systems. Operators of complex systems must process vast amounts of information and suggestions from various devices, subsystems, and both internal and external sources. In addition, they receive requests from multiple entities. This overwhelming influx of data and demands places significant pressure on operators to evaluate and respond swiftly and accurately, which is often crucial to ensuring smooth operation of the entire transport system. To assist operators in making better decisions, new approaches are being introduced, such as expert systems and artificial intelligence. These tools aim to enhance decision-making not only during crises but also for routine operations and more complex tasks related to controlling and monitoring transport systems. The article outlines components of an expert system that uses fuzzy logic to address the complexities of acquiring certain data, particularly from predictive maintenance, which cannot be easily interpreted through simple operational interventions by the operator. Predictive maintenance also relies on decision-making supported by advanced algorithms, which are integrated with the systems technology and control framework.

Air pollution in cities and the possibilities of reducing this pollution represent one of the most important factors that today’s society has to deal with. This paper focuses on a systemic approach to traffic emissions with their relation to meteorological conditions, analyzing the effect of weather on the quantity and dispersion of traffic emissions in a city. Using fuzzy inference systems (FIS) the model for predicting changes in emissions depending on various conditions is developed. The proposed model is based on traffic, meteorology and emission data measured in Prague, Czech Republic. The main objective of the work is to provide insight into how urban planners and policymakers can plan and manage urban transportation more efficiently with environmental protection in mind.

Software quality assurance relies heavily on software reliability as one of its primary metrics. Numerous studies have been conducted to identify the software reliability. Improved software dependability may be studied using a triangular approach that includes software modeling, measurement, and improvement. Each of these steps is critical to the development of a solid software system. Improved accuracy in calculating dependability is critical to managing the quality of software. It has been discovered that deep learning algorithms are excellent methods of assessing many aspects of software dependability. Software systems contain distinct characteristics that can be addressed using deep learning techniques. In this study, a deep-learning-based bidirectional attention-based Zeiler-Fergus convolutional neural network (BA-ZFCNN) technique has been suggested to assess software dependability. In the beginning, the data were standardized by using the scalable error splash method. This approach was then used to extract the software fault-related characteristics using hypertuned evolutionary salp swarm optimization (HESSO). Finally, the Zeiler-Fergus convolutional neural network based on bidirectional attention (BA-ZFCNN) may be used to assess software dependability. The suggested method is used to forecast how many defects or failures there are in a software product. AR1 software defect data is widely used to test the effectiveness of deep learning and traditional machine learning methods. The experimental results reveal that the proposed methods accuracy (96.7%) is higher than the current techniques accuracy.

Global Navigation Satellite Systems are a critical positioning, navigation, and timing source for various industries. However, their weak signal on Earth’s surface makes them vulnerable to jamming. This paper explores the use of machine learning image recognition for categorizing GNSS jamming signals. The study uses data from a long-term monitoring campaign, with over 2,000 jamming events recorded. Seven commonly used jamming signal types were analyzed using the Residual Neural Networks (ResNet). Five different ResNet models with 18 to 152 layers were evaluated, with the best performing achieving a precision greater than 90% in determining the correct jamming signal category.