In high-risk systems, nonconformities are frequently managed as items to be closed rather than as evidence that safety requirements, barriers, or controls may require revision. This study develops and illustrates a control-based framework for translating nonconformities, investigation findings, and safety recommendations into reconstructed safety requirements across the system life cycle. A structured public-source dataset contains 109 retained coded observations from 31 official investigations issued by six agencies in four countries. The unit of analysis is a documented finding and its analytical translation into a revised or revisable requirement. Targeted source-content checks distinguish full reports from official synopses and summaries. Operation was the most frequent affected life-cycle phase (41 observations, 37.6
Sustainable design research has evolved from a primary emphasis on environmental efficiency toward the integration of human-centered dimensions, particularly ergonomics and occupational safety. Existing studies remain fragmented, limiting the development of integrated systems capable of simultaneously ensuring safety, resilience, and long-term sustainability. This gap is especially pronounced in labor-intensive sectors in developing countries, where high physical workloads, informal work practices, and limited knowledge management systems remain prevalent. This study proposes a novel Design Safety Framework that integrates ergonomic risk assessment, design longevity, safety behavior, servant leadership and cultural identity into a unified decision-support model for designing labor-intensive agricultural systems. Unlike previous studies that evaluate these aspects separately, the proposed framework explicitly links workers’ physiological limitations with sustainable engineering design to support safer and more resilient solar water pumping systems. The framework was developed using a sequential, explanatory mixed-methods approach to evaluate the interaction between ergonomic risk, design longevity, and safety behaviour among farmers in Nganjuk, East Java, Indonesia. Data were collected from 45 farmers engaged in repetitive manual agricultural activities, and the relationships among constructs were analysed using Partial Least Squares Structural Equation Modeling (PLS-SEM). The subsequent product development stage focused on enhancing the portability and compactness of the solar-powered water pumping system. The redesigned system incorporates smaller solar panels that can be securely accommodated within dedicated hard cases, making the overall system easier to transport, deploy, and store. The prototype is organized into four dedicated hard cases, each serving a specific functional module, to facilitate transportation, field deployment, operation, and maintenance. The results indicate that H1. Ergonomic Risk affect Safety Behaviour, H3. Design Longevity affect Safety Behaviour, and H7. Servant Leadership affect Safety Behaviour were supported, confirming that Ergonomic Risk, Design Longevity, and Servant Leadership have significant positive effects on Safety Behaviour, respectively. These results suggest that sustainability in labor-intensive systems cannot be achieved solely through efficiency-oriented approaches but must also account for human physiological limitations and adaptive safety considerations. The proposed framework contributes to the development of safer and more sustainable solar water pumping systems, both theoretically and practically, particularly within developing-country agricultural contexts.
This paper introduces a novel and versatile continuous probability distribution, termed the MEGA distribution, as an alternative to the one-parameter Lindley distribution. The proposed model is constructed through a distinctive weighting mechanism in which the gamma component is assigned greater weight than the exponential component, enhancing modeling flexibility. Interestingly, the weights of the MEGA and Lindley distributions intersect at the golden ratio, revealing a unique mathematical connection between the two models. We derive several important distributional properties of the MEGA distribution. The model is shown to exhibit unimodality, a decreasing mean, variance, and mean residual life function, along with increasing coefficient of variation, skewness, kurtosis, and hazard rate function. Furthermore, a transformation relationship between the MEGA and Lindley distributions is developed. Parameter estimation is discussed using both the maximum likelihood method and the method of moments. Simulation procedures for generating MEGA random variates are presented. To demonstrate its practical applicability, the performance of the MEGA distribution is evaluated using two real-world datasets. Overall, the MEGA distribution provides substantial theoretical depth, practical modeling advantages, and promising avenues for future research in reliability theory, survival analysis, and risk modeling.
This study investigates the uncertainty measures of extropy (Ex) and weighted extropy (WEx) for the generalized inverse exponential distribution (GIED), particularly when observations are limited to upper record values. Despite the widespread use of GIED in reliability and survival analysis, its Ex-based uncertainty measures remain largely unexplored, and WEx for GIED has not been previously studied. Our research fills this gap by deriving analytical expressions for both Ex and WEx within the GIED framework. To estimate these measures, we employ maximum likelihood and Bayesian techniques. We consider both non-informative and informative priors for Bayesian estimation and evaluate estimates under symmetric (squared-error) and asymmetric (weighted-squared-error, minimum expected) loss functions. Given the complexity of the Bayesian estimators, a Markov Chain Monte Carlo approach is adopted for computation. Furthermore, the study constructs bootstrap confidence intervals and highest posterior density intervals to evaluate the performance of the estimators. The efficacy of the proposed methods will be rigorously assessed through simulation studies, comparing mean squared errors, absolute biases, average confidence interval lengths, and coverage probabilities. Finally, the practical utility of our findings is demonstrated through the analysis of two real-world datasets in reliability engineering and maintenance analysis. This work enhances our understanding of GIED’s uncertainty characteristics under record-value schemes, providing valuable tools for modeling and prediction across diverse fields.
This paper recommends a Multiple Dependent State Repetitive (MDSR) sampling plan for attributes sampling considering the same Inverted NadarajahHaghighi (INH) distribution. Though there are several dependent and repetitive state sampling plans developed under various lifetime distributions there was limited study on MDSR sampling plans based on percentile lifetimes under the INH distribution. To fill this, a new MDSR sampling plan is created based on percentile ratio as a quality indicator. The two-point approach is used to find the best value of the design parameter that minimises the producer and consumer risk levels, giving a set of operating characteristics (OC) values. Performance of the proposed plan is tested for various shape parameters, various risks faced by the consumers and various percentile ratios and compared by using the minimum Average Sample Number (ASN) between them. The results indicate that the MDSR plan would typically take significantly fewer observations than the two existing Multiple Dependent State (MDS) and Single Sampling Plan (SSP) plans. An example of this can be seen for the case of percentile ratio 2.0 and consumers risk of 0.10, where the proposed MDSR plan offers a better ASN (41.30) than the MDS plan (45) and the SSP (73) when (a = 0.5). Likewise, for the value (a = 1.0), the ASN is 38.76 whereas for the MDS and SSP the ASN are 43 and 68, respectively. Under the parameter settings considered, the proposed approach is more efficient in terms of sampling for the given producers and consumers risks. The usefulness of the suggested MDSR sampling plan applied to a real-life data application—quality-control/reliability management, is further illustrated. The methodology proposed in the present document in turn offers a low cost and efficient sampling methodology that can be used in products with a distributed lifetime distribution similar to INH.
