Railways remain an essential part of modern transportation, yet their safe functioning is often determined by the actual condition of the tracks. The study looks at various ways to detect faults in the rail infrastructure and splits them broadly into two categories: static and dynamic techniques. Different countries rely on different tools to monitor tracks, and this paper compares those tools based on practical factors like how precise they are, how much ground they cover, and how difficult or costly they are to operate. Rather than simply listing pros and cons, we try to show where each method works best. To make sense of the data, visuals like charts and summaries were added, making it easier to see where each approach fits. One part of the analysis pays special attention to how certain features of the railway - such as how wide the rails are or how much the outer rail is elevated - can influence the choice of inspection methods. Lately, there's been a shift toward smarter diagnostics. Technologies like AI, digital simulations of tracks (known as digital twins), and systems using many sensors at once are gaining ground. These tools are changing how track inspections are done and offer new opportunities for early problem detection. This paper doesn't just list methods - it gives a clear structure for understanding which approach fits what context. The outcomes can help transportation teams fine-tune how they take care of tracks and make the system more dependable in the long term.

This paper presents and experimentally validates an electronic voting model that stores results in a decentralized blockchain while using biometric voter verification and automatic detection of synthetically generated (deepfake) facial images. For identity verification, Vision Transformer and ResNet-50 architectures are fine-tuned with LowRank Adaptation (LoRA). Frequency-domain information from the Fast Fourier Transform (FFT) is also included as a descriptor for detecting synthetic images. Since blockchain records cannot be altered, the model introduces a controlled revoting process based on self-destructing ballots and timestamps recorded on the blockchain; the votes themselves are submitted through an Ethereum smart contract. In the experiments, the LoRA-adapted ViT configuration distinguishes real from synthetic facial images with an accuracy of 97.48%, exceeding the performance of the LoRA-adapted ResNet-50 baseline. However, because no separate FFT ablation study has been conducted, this work does not claim that the frequency descriptor makes an independent contribution. The proposed system is presented as part of Smart City digital infrastructure and lays the groundwork for future research on privacypreserving biometric protocols and blockchain platforms capable of scaling to electoral processes.

We perform a cross-layer penetration testing on one of the most popular Wi-Fi smart locks (Tuya 902V). The methodology combines wireless traffic analysis using an Alfa AWUS036AXML adapter, forced re-association via deauthentication to make Wi-Fi Protected Access 2 (WPA2) 4-way Extensible Authentication Protocol over LAN (EAPOL) handshake visible with Airodump/Aireplay, offline dictionary attack with Aircrack-ng, Android app reverse engineering using Apktool, Jadx, and MobSF; denial-of-service experiment (DoS) executed by hping3; Near-Field Communications (NFC)/Radio-Frequency Identification (RFID) key-clone attempt by Flipper Zero. Handshake is empirically captured but no Wi-Fi passphrase found under 14M dictionary entries; DoS test breaks cloud notification and remote control features; MObsF-assisted analysis shows dangerous permissions, many exported components, sensitive data being logged and hardcoded strings even though communication between app and lock is encrypted; Registered MIFARE Classic 1K credential fully read (including both A/B keys) however not able to emulate because of Flipper Unique Identifier (UID)/emulation limitations. This work tackles the issue of having no systematic or experimental method for discovering vulnerabilities in smart lock systems across several layers. We need to investigate how various attacks, for instance, via wireless, reverse-engineering applications, through Denial of Service, or cloning of credentials, could impact the overall security of commercially available products. We summarize the mitigations and check their relevance: enable Protected Management Frames (802.11 w) and move to Wi-Fi Protected Access 3—Simultaneous Authentication of Equals (WPA3-SAE) to reduce de-auth/handshake usefulness, enforce strong Pre-Shared Keys (PSKs), use TLS with cert pinning, rate-limit and make device APIs fault-tolerant, enforce signed/anti-rollback firmware updates, and replace static RFID tokens with Data Encryption Standard Fast Innovative Reliable and Secure (DESFire)/rolling-code credentials. The results show real attack surfaces across the network and application layers and provide actionable hardening guidance for smart-lock vendors and operators.

The Internet has changed dramatically over the past decade, and search engines have worked hard to adapt to these changes by providing better content to users. Many efforts have been made to assess document quality, including attempts to use semantic analysis to extract meaningful information from material presented on the Internet. This article describes the development of a text editor that uses morphological and syntactic analyzers to process website content. The main purpose of the system is to automatically mark and improve the text content of web pages by integrating semantic information obtained based on text structure and meaning analysis. The system uses preprocessing technology and deep NLP models (such as BERT), combined with a self-built Kazakh tree library (ConLL-U format) to restore the query's lemmatization. Optimize the matching of queries and documents through TF-IDF, Word2Vec and query expansion technology. Experiments were conducted on an autonomously collected Kazakh news dataset. Experimental results show that morphological analysis significantly improves retrieval accuracy during the text normalization process, while the deep learning model further enhances the ranking effect based on semantic correlation. This research provides new technical means for content optimization and information retrieval in low-resource languages, and helps improve the performance of search engines in multi-language environments.

In this paper, we present a review of the latest deep learning methods used in inspecting railway infrastructure with vision systems. Traditional methods for inspecting railways are not always efficient, scalable or accurate, so using computer vision and artificial intelligence is now needed to maintain rail safety. The research divides more than 200 papers into important themes, covering rail surface fault detection, issues with fasteners and sleepers, adding synthetic data, using different types of railroad sites and multi-technology integration. For every domain, we cover current progressive models, for example, CNNs, YOLOv5, U-Net, Faster R-CNN and Vision Transformers, discussing their uses, pros and cons. The problem of using the model in practice is analyzed by studying generalization, instant results, understanding how it learns and the hardware limitations. We also find areas where progress is needed and suggest future solutions to reduce the difference between what academic discovery and what industries can use.

Electronic voting requires the simultaneous admission of only legitimate participants, ballot uniqueness, vote confidentiality, storage integrity, and result verifiability. Blockchain alone does not solve these problems, since ledger immutability does not guarantee anonymity, ballot correctness, or reduced trust concentration. The purpose of this work is to develop a parameterizable research framework for electronic voting scenarios with enhanced cryptographic protection, allowing the security level to be varied according to the requirements of a voting scenario. The main contribution of the work is a parameterizable research architecture for composing and experimentally comparing electronic voting configurations with different security and computational profiles. The cryptographic and audit mechanisms integrated into this architecture include blind-signature-based anonymous authorization, encrypted ballot submission, blockchain-style audit, receipt verification, homomorphic tally publication, and threshold-supported tally artifacts. These mechanisms are not proposed as new cryptographic primitives; rather, they are integrated into a reproducible prototype to study how their combination affects verifiability, privacy support, auditability, and computational cost. Compared with basic blockchain-based voting prototypes, this architecture explicitly separates security, privacy, and verifiability profiles and makes their computational cost observable. The implemented prototype is used as an experimental platform for analyzing supported security properties, threat modeling, and computational cost estimation. The results show that authentication, anonymous token issuance, and receipt verification maintain an almost constant cost at the studied scale, while the main cryptographic burden is associated with encrypted ballot submission and threshold-supported tally publication. The scientific novelty of the work lies in constructing a parameterizable architecture that integrates several cryptographic mechanisms and a blockchain audit layer into one reproducible research prototype. At the same time, the proposed approach retains prototype-level limitations associated with the absence of a full zero-knowledge proof stack, independently deployed threshold authorities, and coercion-resistance mechanisms.