International Journal of Reliability and Safety (15 papers in press)

Regular Issues

• Open Access

  Intelligent decision support in ideological and political education: a knowledge graph and affective computing approach for organisational reliability and psychological risk management ( Free Full-text Access )

  by Xiaofei Xie 
  Abstract: Ensuring the reliability and safety of higher education institutions requires effective management of student psychological states and behavioural norms. To address weak risk-informed decision support and insufficient recognition of psychological risk indicators, this paper proposes a data-driven knowledge reasoning model integrating GCN-TransE and BiLSTM-Attention. TransE constructs a knowledge graph of organisational norms and behavioural entities, while GCN captures multi-hop relational dependencies for risk factor analysis. BiLSTM-Attention models temporal emotional features for dynamic psychological monitoring. Cross-modal consistency constraints then fuse semantic and emotional representations to generate structured intervention strategies for educational management systems. Experimental results show that relational reasoning accuracy reaches 0.92, emotion recognition F1 reaches 0.914, and the joint representation consistency index reaches 0.87. The proposed framework provides an effective technical path for intelligent risk management and human-centred decision support in higher education.
  Keywords: intelligent risk management; organisational reliability; affective computing; GCN-TransE; knowledge graphs; decision support systems; human factors in safety.
  DOI: 10.1504/IJRS.2026.10077842
   
  

• Open Access

  Bayesian system dynamics for dynamic risk management and reliability-oriented resource performance in sustainable environmental systems ( Free Full-text Access )

  by Li Shanshan 
  Abstract: Uncertainty management and economic efficiency are critical challenges in intelligent and sustainable environmental design due to the absence of a unified modelling framework capable of capturing complex feedback mechanisms. To address these limitations in system reliability and decision-making, this study proposes a data-driven Bayesian system dynamics modelling approach. The framework employs system dynamics to construct causal-loop and stockflow structures that represent the interactions among energy, resource, and behavioural subsystems within complex socio-technical environmental systems. Integrated with a Bayesian inference-based parameter updating mechanism, the model utilises Markov Chain Monte Carlo (MCMC) sampling to iteratively refine posterior parameter distributions. This forms a closed-loop process of dynamic risk assessment and parameter updating, enabling accurate failure prediction and robust managerial decision support. Experimental results demonstrate that the proposed approach achieves a prediction accuracy of 92.4%, improves system-level resource utilisation performance with an operational utilisation rate of 87.6%, and enhances system reliability with a long-term stability index of 91.2% under complex environmental scenarios. The findings validate the effectiveness of the proposed method for intelligent risk mitigation and cost-effective environmental management, offering both theoretical insight and methodological support for reliability-centred performance regulation and long-horizon system stability management. Keywords: Bayesian system dynamics; sust
  Keywords: Bayesian system dynamics; sustainable environmental design; intelligent environmental modeling; parameter adaptive inference; multi-level system optimization.
  DOI: 10.1504/IJRS.2026.10077985
   
  

• Open Access

  AI-enabled sustainable decision support for large-scale sports facility operations: integrating ASTGCN forecasting and knowledge graph reasoning ( Free Full-text Access )

  by Nan Gao, Zi-Yuan Wang, Kai Zhou, Run-Hang Chen 
  Abstract: Large-scale sports facilities function as complex socio-technical systems in which unreliable environmental control and misaligned energy allocation introduce operational risks, degrade comfort reliability, and increase economic losses. Addressing these challenges, this study develops an AI-enabled, risk-aware decision support system that integrates short-term load prediction with interpretable knowledge-based reasoning to achieve economically efficient and reliable air-conditioning management under uncertainty. Multivariate data from 12 air-conditioning zones are modelled using an Attention Spatio-Temporal Graph Convolutional Network (ASTGCN), which achieves RMSE and MAPE values of 7.1 kWh and 9.2% and provides dependable forecasts for proactive control. An application specific knowledge graph is then constructed, where semantic mappings and SWRL rules translate ASTGCN outputs into logical facts to generate transparent, risk-informed operational strategies. A 61-day evaluation shows that the integrated framework reduces energy consumption by 15.9% while limiting comfort deviation to 4.2%, demonstrating simultaneous improvement in economic performance, operational reliability and sustainable facility operation. The results confirm the value of combining predictive machine learning with interpretable cognitive reasoning for intelligent risk management and decision-making in complex infrastructure systems.
  Keywords: sports facility operation; risk-aware decision support system; reliability and comfort management; energy efficiency and economic performance; ASTGCN; knowledge graph; SWRL reasoning; intelligent HVAC control; socio-technical systems.
  DOI: 10.1504/IJRS.2026.10078077
   
  
• Pre-chamber spark ignition: a reliability analysis of pre-chamber valve functions  
  by Faraz Akbar, Sarah Zaki 
  Abstract: A pre-chamber ignition allows spark-ignition engines to operate in lean air-fuel settings. It improves fuel efficiency and reduces emissions. In this study, a reliability analysis of a single GE Jenbacher J620 natural gas engine was done. It was operational on continuous load in the power generation sector in Karachi, Pakistan. A bathtub curve of the GE J620 pre-chamber gas valve (PCV) was generated. The three-year industrial data comprised PCV failures that occurred between two overhauls. During infant mortality, the curve revealed 7 failures during 1000 hours. This decreased to a failure for the next two cycles of thousand hours each. There was a 40% decrease in reliability after 1500 hours. Exponential distribution revealed that the mean time-to-failure (MTTF) was 545.5 hours. This study was the first of its kind in the facility. Previously, much time was lost in breakdown maintenance. Thus, it helped to increase the systems reliability.
  Keywords: bathtub curve; exponential distribution; failure rate; fuel injection; gas engine; pre-chamber combustion; pre-chamber spark ignition; pre-chamber valve; probability density function; reliability.
  
  
• A comparative analysis of neural network and ensemble learning models for automated root cause classification in fatal mine accident reports  
  by Kumar Arra, Suprakash Gupta 
  Abstract: Mining is one of the highest-risk industries due to the hazardous nature of coal extraction, a primary power generation source. This study uses 1305 mining accident reports, recorded from 1995 to 2015 in India, to determine the root causes of accidents and devise effective safety measures. The methodology follows a structured approach combining expert domain knowledge with advanced natural language processing techniques. We employed a novel hybrid vectorisation strategy combining N-grams, TF-IDF, and word embeddings to analyse accident reports. We stratified a 10-fold cross-validation used to address the imbalanced data distribution. This study compared three machine learning models: CatBoost, XGBoost and neural networks. CatBoost demonstrated superior performance with a 0.91 F1-score and 0.98 PR-AUC, outperforming both XGBoost (0.88, 0.96) and neural networks (0.84, 0.93). The developed system predicts accident root causes with 91% accuracy, providing a robust framework for improved decision-making and enhanced industry safety standards.
  Keywords: automated classification; accident prevention; ensemble learning; machine learning; mine accidents; neural networks; root cause analysis; text mining.
  DOI: 10.1504/IJRS.2025.10074662
   
  
• Automatic track monitoring and fault detection using vibration sensor data in railway transport system  
  by Si Chen 
  Abstract: The development of a real-time automatic track monitoring system can detect the condition of the track without interfering normal traffic. The objective of this study is to classify railway tracks as healthy or defective with the help of vibration readings and environmental factors such as temperature and humidity. The proposed fault detection system investigates the efficiency of several ML algorithms, which are combined with the Subtraction-Average-Based Optimizer for hyperparameter tuning to improve their performance in classifying railway tracks. According to results, the SABO-RF model provides a strong and reliable approach to real-time fault detection with 99.86% accuracy and 99.84% precision, which contributes toward preventing accidents and minimizing operational disruption in railway systems. The sensitivity of the SABO-RF model outlines the influence of vibration in dimension y with a strong positive contribution of +4.24 and a strong negative contribution of -3.87 in classifying railway tracks by model.
  Keywords: railway transportation system; railway track; fault detection; vibration sensors data; data-driven approach; hyperparameter tuning.
  DOI: 10.1504/IJRS.2025.10075056
   
  
• Rolling bearing fault vibration characteristic analysis software system accounting waviness and local defects  
  by Wujiu Pan, Yuanbin Chen, Conghui Han, Junyi Wang, Jianwen Bao 
  Abstract: As the core component of rotating machinery, the health status of rolling bearing directly affects the reliability and safety of equipment operation. Aiming at the vibration characteristics caused by bearing waviness and local defects, a dynamic model considering the waviness parameters of inner and outer races is established, and a time-varying displacement excitation local defect model based on piecewise function is proposed. The displacement, velocity, acceleration and frequency domain response characteristics of the bearing are accurately analyzed through nonlinear dynamic equations. On this basis, a software system integrating waviness analysis, single fault and composite fault simulation is developed based on MATLAB App Designer. The results show that the proposed model can effectively identify the waviness characteristic frequency (such as ball passing frequency and its harmonics) and local defect impact signals, and provide theoretical support and practical tools for bearing fault diagnosis and health monitoring in complex industrial scenes.
  Keywords: Rolling bearing;Waviness;Local defects;Composite fault diagnosis Software system development.
  
  
• Importance of component states under reliability correlation of multi-state systems based on multi-valued structure function  
  by Emad K. Mutar, Zahir Abdul Haddi Hassan 
  Abstract: The structure function is crucial for multi-state systems (MSSs), but its complexity rises as the number of components and states increases, making management difficult. It is important to address these challenges and formulate techniques to ensure the reliability of MSSs. This paper presents a new technique for assessing MSS reliability using multi-valued structure function and multi-valued decision diagram (MDD). It assumes component states follow a binomial distribution and explores the effects of the correlation coefficient. The paper evaluates Birnbaum Importance (BI), Risk Achievement Worth (RAW), and Risk Reduction Worth (RRW) for each component in the MSS across various performance levels. It simplifies computations by using subsystems, the chain rule, and conditional probability. The methods are demonstrated with numerical examples and an application related to a spacecraft system, showing how these measures improve reliability and confirm the effectiveness of the techniques in assessing reliability and component importance.
  Keywords: structure function; multi-state system; system performance; multi-valued decision diagram; component importance.
  DOI: 10.1504/IJRS.2025.10075196
   
  
• Reliability analysis of reconfigurable integrated modular avionics: a modelling-based approach  
  by Changxiao Zhao, Daojun Li, Yi Tian 
  Abstract: With the escalating complexity of flight environments and the growing sophistication of avionics capabilities, dynamic reconfiguration has emerged as a critical enabling technology for modern Integrated Modular Avionics (IMA) systems. While this reconfigurability enhances operational adaptability, it introduces predictive uncertainties that compromise system reliability and pose significant challenges to aviation safety. This paper presents a novel reliability assessment framework for reconfigurable IMA systems through three key contributions. The results show that compared with the traditional model, the proposed method significantly improves the accuracy of reliability evaluation, and the system reliability is increased by 38.02%. The proposed framework makes up for the critical defect of uncertainty perceived reliability analysis and provides a reference for optimizing reconfiguration strategies in the early development stage of safety-critical avionics systems.
  Keywords: complex avionics system architecture; safety-critical system; dynamic reconfiguration mechanism; reliability analysis.
  DOI: 10.1504/IJRS.2025.10075458
   
  
• Dynamic characteristics of the neutron pulsing chopper rotor system based on the finite element method and the response surface method  
  by Liangwei Zhang, Rongchang Zhang, Qing Zhang, Weiliang Cai, Ping Wang, Zhengmin Hu 
  Abstract: The high-speed operation of neutron pulse choppers often results in system instability caused by resonance, compromising reliability and safety. To address this, we conducted an in-depth study of the rotor dynamics of neutron pulse choppers. A finite element model of the rotor system was developed to calculate its critical dynamic characteristics. Using central composite design and response surface methodology, we analyzed the effects of bearing stiffness and position on critical speeds. Results show that the second critical speed is close to the operating speed, posing a resonance risk, with high vibration amplitudes observed near the chopper disk and the shaft’s front end. Bearing stiffness positively correlates with critical speeds, showing the highest sensitivity for the first critical speed. The front and rear bearings significantly influence the second and first critical speeds, respectively. This study provides theoretical guidance for optimizing high-performance rotors in neutron pulse choppers and similar high-speed systems.
  Keywords: rotor dynamics; neutron pulse chopper; modal analysis; critical speed; response surface methodology.
  DOI: 10.1504/IJRS.2025.10075461
   
  
• Co-enhancement model of multiple ultra-high voltage converter stations resilience under seismic impact  
  by Guo Li, Xue Zhihang 
  Abstract: Ultra-high voltage converter stations are one of the most critical power infrastructures to maintain stable and efficient power transmission over a long distance. Enhancing the resilience of converter stations against earthquakes becomes a prior task in modern power systems. In this article, considering the investment budget and space limitation for spare equipment in a converter station, a cooperative dispatching framework for multiple converter stations is proposed to achieve resilience enhancement. Driven by minimizing the performance loss of converter stations subject to seismic impact, the co-enhancement model is dedicated to hardening equipment, spare parts allocation, and the sequential repair scheme. Compared with the individual decision for each converter station in the traditional model, the emergency resources cooperative dispatching among multiple converter stations is considered in this model to achieve the mutual coordination of the widlely available resources. The real-world ultra-high voltage converter stations in southwest China are chosen to validate the efficacy of the proposed model. The results demonstrate that the co-enhancement model can improve earthquake resilience through collaborative dispatching among multiple converter stations.
  Keywords: ultra-high voltage converter stations; seismic impact; co-enhancement model; mutual coordination.
  DOI: 10.1504/IJRS.2025.10076147
   
  
• Enhancing emergency operations in urban railways under waterlogging: a PERT-BN based time-risk analysis approach  
  by Jinduo Xing, Jiagi Qian, Tao Tang 
  Abstract: Unexpected incidents or accidents during urban railways operations, such as waterlogging, may lead to catastrophic consequences even casualties Effective emergency operations are designed to ensure the continuity and safety operation of the urban railway against disruptive events It is highly required to ensure the efficiency and safety of emergency operations In this paper, we proposed an integrated quantitative time-risk analysis method of emergency operations to improve the efficiency and safety of emergency processes The Program Evaluation and Review Technique (PERT) model is employed to quantitatively guide emergency operations Moreover, the critical path and the total duration time for urban railway emergencies are determined Then, the associated risk in the critical path is analyzed resorting to Bayesian networks (BN) Finally, the proposed method is exemplified on a real urban railway waterlogging accident The results demonstrate the feasibility and effectiveness of the developed PERT-BN model.
  Keywords: urban railway emergency; waterlogging; program evaluation and review technique; Bayesian networks.
  DOI: 10.1504/IJRS.2025.10076423
   
  
• A systematic review on safety and security issues of the autonomous vehicles  
  by Ali Mahmood, Robert Szabolcsi 
  Abstract: Autonomous vehicles (AVs) are transforming modern transportation by improving road safety, reducing traffic congestion, and enhancing mobility. However, safety and security remain major challenges due to the integration of artificial intelligence, sensor fusion, machine learning, and vehicle-toeverything (V2X) communication. This systematic review analyses safety and security issues in AVs by synthesising findings from 80 academic papers across multiple disciplines, including computer science, transportation engineering, and control systems. The review focuses on five key areas influencing AV deployment: collision avoidance strategies, cybersecurity threats, human interaction and perception, regulatory and ethical considerations, and testing and validation methods. It discusses approaches such as model predictive control and reinforcement learning for collision avoidance, secure V2X communication for cybersecurity, and simulation-based and real-world testing for validation. The study highlights current challenges in AV safety, regulation, and public trust, and identifies future research directions including hybrid AI-driven safety mechanisms, standardised global regulations, and digital-twin-based validation frameworks.
  Keywords: autonomous vehicles; safety and security; collision avoidance; cybersecurity; human-AV interaction; AV ethics; testing and validation.
  DOI: 10.1504/IJRS.2026.10077218
   
  
• Dynamic detection of emergencies based on a hybrid architecture of multi-view clustering and multilayer perceptron  
  by Yameng Bai, Junxia Meng, Shuai Zhao, Ruoyu Ren 
  Abstract: To improve the dynamic detection of emergencies, this paper proposes a method based on multi-view clustering and an improved Multilayer Perceptron (MLP). It integrates social media text, sensor data, and geographic location information. An enhanced Multi-View Spectral Clustering (MVSC) algorithm with feature weight learning and adaptive view fusion is used for pre-processing. The improved MVSC achieves a clustering accuracy of 0.917 and a Davies-Bouldin Index of 0.45, converging in 80 iterations. For detection, an MLP incorporating a multi-head attention mechanism is developed. The proposed model achieves the highest detection rates across different emergency types: 0.927 for natural disasters, 0.912 for social events, and 0.935 for public health events, with response times of 0.24s, 0.27s, and 0.22s, respectively. Results demonstrate that the hybrid architecture enables fast and accurate identification of emergencies, contributing to improved emergency response and reduced socio-economic impacts.
  Keywords: dynamic detection of emergencies; MVSC; MLP; multi-head attention mechanism; hybrid architecture.
  DOI: 10.1504/IJRS.2026.10077986
   
  
• Exploring the impact of different safety training methods on cognitive engagement: an EEG study in chemical plants  
  by Yuyao Deng, Mingyue Zhu, Zhengwen Zhou, Yong Wan, Hailin Guo 
  Abstract: Safety training enhances cognitive frameworks in chemical plants, reducing errors and risks during emergencies. However, bureaucratic and subjective factors often limit its efficacy. With Safety Education 4.0, novel methods like virtual reality and web-based tools are emerging, yet trainees physiological responses remain understudied. This research in Hubei Province surveyed 20 chemical enterprises and conducted EEG experiments (via 32-channel Neuroscan) with 25 participants exposed to six stimuli types. Data analysis focused on Kirkpatrick model indicators (Interest, Attractiveness, Attention, Memorisation) using paired t-tests. Results showed pre-recorded audio (RSM) significantly boosted Interest, Attractiveness, and Attention, while audiovisual subtitled stimuli (OVS) optimised Memorisation. Findings highlight tailored training media selection to balance engagement and knowledge retention, aiding cognitive load reduction for safety personnel. This study underscores EEGs potential in objectively evaluating training impacts within industrial education frameworks.
  Keywords: safety training methods; Kirkpatrick model; EEG; electroencephalogram; chemical safety; absolute power; audio-visual stimulation.
  DOI: 10.1504/IJRS.2026.10078078