Industrial and occupational ergonomics in the petrochemical process industry: a regression trees approach

Establishment-level safety analytics: a scoping review

The use of data analytics has seen widespread application in fields such as medicine and supply chain management, but their application in occupational safety has only recently become more common. The purpose of this scoping review was to summarize studies that employed analytics within establishments to reveal insights about work-related injuries or fatalities. Over 300 articles were reviewed to survey the objectives, scope and methods used in this emerging field. We conclude that the promise of analytics for providing actionable insights to address occupational safety concerns is still in its infancy. Our review shows that most articles were focused on method development and validation, including studies that tested novel methods or compared the utility of multiple methods. Many of the studies cited various challenges in overcoming barriers caused by inadequate or inefficient technical infrastructures and unsupportive data cultures that threaten the accuracy and quality of insights revealed by the analytics.

Prediction of geomechanical bearing capacity using autoregressive deep neural network in carbon capture and storage systems

Carbon Capture and Storage (CCS) field is growing rapidly as a means to mitigate the accumulation of greenhouse gas emissions. However, the geomechanical stability of CCS systems, particularly related to bearing capacity, remains a critical challenge that requires accurate prediction models. In this research paper, we investigate the efficacy of employing an Autoregressive Deep Neural Network (ARDNN) algorithm to predict the geomechanical bearing capacity in CCS systems through shear wave velocity prediction as an index for bearing capacity evaluation of deep rock formations. The model utilizes a dataset consisting of 23,000 data points to train and test the ARDNN algorithm. Its scalability, use of deep learning techniques, automatic feature extraction, adaptability to changes in data, and versatility in various prediction tasks make it an attractive option for accurate predictions. The results demonstrate exceptional performance, as evidenced by an R-squared value of 0.9906 and a mean squared error of 0.0438 for the test data compared to the measured data. This research has significant practical implications for effectively predicting geomechanical stability in CCS systems, thus mitigating potential risks associated with their operation.

Characteristics of fatal occupational injuries in migrant workers in South Korea: A machine learning study

Objective

Analysis of occupational injuries is essential for developing preventive strategies. However, few studies have evaluated severe occupational injuries in migrant workers from the perspective of gender. Therefore, using a new analytical method, this study was performed to identify gender-specific characteristics associated with fatal occupational injuries among migrant workers; the interactions between these factors, were also analyzed. In addition, we compared the utility of explainable artificial intelligence (XAI) using SHapley Additive exPlanations (SHAP) with logistic regression (LR) and discuss caveats regarding its use.

Materials and methods

We analyzed national statistics for occupational injuries among migrant workers (n = 67,576) in South Korea between January 1, 2007, and September 30, 2018. We applied an extreme gradient boosting model and developed SHAP and LR models for comparison.

Results

We found clear gender differences in fatal occupational injuries among migrant workers, with males in the same occupation having a higher risk of death than females. These gender differences suggest the need for gender-specific occupational injury prevention interventions for migrant workers to reduce the mortality rate. Occupation was a significant predictor of death among female migrant workers only, with care jobs having the highest fatality risk. The occupational fatality risk of female workers would not have been identified without the performance of detailed job-specific analyses stratified by gender. The major advantages of SHAP identified in the present study were the automatic identification and analysis of interactions, ability to determine the relative contributions of each feature, and high overall performance. The major caveat when using SHAP is that causality cannot be established.

Conclusion

Detailed job-specific analyses stratified by gender, and interventions considering the gender of migrant workers, are necessary to reduce occupational fatality rates. The XAI approach should be considered as a complementary analytical method for epidemiological studies, as it overcomes the limitations of traditional statistical analyses.

Classification and pattern extraction of incidents: a deep learning-based approach

Classifying or predicting occupational incidents using both structured and unstructured (text) data are an unexplored area of research. Unstructured texts, i.e., incident narratives are often unutilized or underutilized. Besides the explicit information, there exist a large amount of hidden information present in a dataset, which cannot be explored by the traditional machine learning (ML) algorithms. There is a scarcity of studies that reveal the use of deep neural networks (DNNs) in the domain of incident prediction, and its parameter optimization for achieving better prediction power. To address these issues, initially, key terms are extracted from the unstructured texts using LDA-based topic modeling. Then, these key terms are added with the predictor categories to form the feature vector, which is further processed for noise reduction and fed to the adaptive moment estimation (ADAM)-based DNN (i.e., ADNN) for classification, as ADAM is superior to GD, SGD, and RMSProp. To evaluate the effectiveness of our proposed method, a comparative study has been conducted using some state-of-the-arts on five benchmark datasets. Moreover, a case study of an integrated steel plant in India has been demonstrated for the validation of the proposed model. Experimental results reveal that ADNN produces superior performance than others in terms of accuracy. Therefore, the present study offers a robust methodological guide that enables us to handle the issues of unstructured data and hidden information for developing a predictive model.

Comparison of machine learning methods in predicting binary and multi-class occupational accident severity

Although Machine Learning (ML) is widely used to examine hidden patterns in complex databases and learn from them to predict future events in many fields, utilization of it for predicting the outcome of occupational accidents is relatively sparse. This study utilized diversified ML algorithms; Multinomial Logistic Regression (MLR), Support Vector Machines (SVM), Single C5.0 Tree (C5), Stochastic Gradient Boosting (SGB), and Neural Network (NN) in classifying the severity of occupational accidents in binary (Fatal/NonFatal) and multi-class (Fatal/Major/Minor) outcomes. Comparison of the performance of models showed Balanced Accuracy to be the best for SVM and SGB methods in 2-Class and SGB in 3-Class. Algorithms performed better at predicting fatal accidents compared to major and minor accidents. Results obtained revealed that, ML unveils factors contributing to severity to better address the corrective actions. Furthermore, taking action related to even some of the most significant factors in complex accidents database with many attributes can prevent majority of severe accidents. Interpretation of most significant factors identified for accident prediction suggest the following corrective measures: taking fall prevention actions, prioritizing workplace inspections based on the number of employees, and supplementing safety actions according to worker’s age and experience.

Predicting Procedure Step Performance From Operator and Text Features: A Critical First Step Toward Machine Learning-Driven Procedure Design

OBJECTIVE

The goal of this study is to assess machine learning for predicting procedure performance from operator and procedure characteristics.

BACKGROUND

Procedures are vital for the performance and safety of high-risk industries. Current procedure design guidelines are insufficient because they rely on subjective assessments and qualitative analyses that struggle to integrate and quantify the diversity of factors that influence procedure performance.

METHOD

We used data from a 25-participant study with four procedures, conducted on a high-fidelity oil extraction simulation to develop logistic regression (LR), random forest (RF), and decision tree (DT) algorithms that predict procedure step performance from operator, step, readability, and natural language processing-based features. Features were filtered using the Boruta approach. The algorithms were trained and optimized with a repeated 10-fold cross-validation. After training, inference was performed using variable importance and partial dependence plots.

RESULTS

The RF, DT, and LR algorithms with all features had an area under the receiver operating characteristic curve (AUC) of 0.78, 0.77, and 0.75, respectively, and significantly outperformed the LR with only operator features (LROP), with an AUC of 0.61. The most important features were experience, familiarity, total words, and character-based metrics. The partial dependence plots showed that steps with fewer words, abbreviations, and characters were correlated with correct step performance.

CONCLUSION

Machine learning algorithms are a promising approach for predicting step-level procedure performance, with acknowledged limitations on interpolating to nonobserved data, and may help guide procedure design after validation with additional data on further tasks.

APPLICATION

After validation, the inferences from these models can be used to generate procedure design alternatives.

Characteristic Analysis of Occupational Confined Space Accidents in Taiwan and Its Prevention Strategy

According to the US Bureau of Labor Statistics (BLS), 882 people were killed or injured due to confined space accidents in 2011–2017. Occupational accident statistics published in 2008–2018 by the Taiwan Occupational Safety and Health Administration (OSHA, Taiwan) show that 70 people suffered from disasters and 52 were injured in the 64 accident reports involving confined spaces. In the US, on average, 126 people die each year in accidents related to confined spaces, and in Taiwan, an average of 8 people per year are casualties of accidents involving confined spaces, proving that it is an area of concern that cannot be neglected. When misjudgments occur in confined spaces, not only can people be hurt, but they can even lose their lives, and the risks associated with confined spaces can subsequently result in rescue personnel also being killed or injured. This study was conducted via the systematic causal analysis technique (SCAT), which was proposed and developed by the International Loss Control Institute (ILCI), with the intention of identifying the critical basic causes of the confined space accidents that have occurred over the years in the Taiwan area, in order to propose corresponding improvement strategies. After investigating the statistics in Taiwan, it was determined that hydrogen sulfide was involved in 45% of accidental deaths in confined spaces, followed by 11% involving carbon dioxide, 9% involving carbon monoxide, and 7% involving toluene. Additional analysis of non-standard acts identified “failure of operating procedures” as being involved in 27% of accidents, followed by 25% involving “improper personal protective equipment” and 23% involving “incorrect position”. The analysis of non-standard conditions revealed that “dangerous workplace” was involved in 39% of accidents, “improper protective measures” in 30%, and “inadequate ventilation” in 27%. In accordance with our analysis results, it could be suggested that hazard prevention strategies for confined spaces, in addition to encouraging avoidance of non-standard acts by personnel, should also strive to improve these non-standard conditions. Otherwise, if not prevented deliberately and in a fundamental, relevant accidents will remain inevitable.

Quantitative Risk Assessment of Seafarers' Nonfatal Injuries Due to Occupational Accidents Based on Bayesian Network Modeling

Reducing the incidence of seafarers' workplace injuries is of great importance to shipping and ship management companies. The objective of this study is to identify the important influencing factors and to build a quantitative model for the injury risk analysis aboard ships, so as to provide a decision support framework for effective injury prevention and management. Most of the previous research on seafarers' occupational accidents either adopts a qualitative approach or applies simple descriptive statistics for analyses. In this study, the advanced method of a Bayesian network (BN) is used for the predictive modeling of seafarer injuries for its interpretative power as well as predictive capacity. The modeling is data driven and based on an extensive empirical survey to collect data on seafarers' working practice and their injury records during the latest tour of duty, which could overcome the limitation of historical injury databases that mostly contain only data about the injured group instead of the entire population. Using the survey data, a BN model was developed consisting of nine major variables, including "PPE availability," "Age," and "Experience" of the seafarers, which were identified to be the most influential risk factors. The model was validated further with several tests through sensitivity analyses and logical axiom test. Finally, implementation of the result toward decision support for safety management in the global shipping industry was discussed.

Segmentation of Severe Occupational Incidents in Agribusiness Industries Using Latent Class Clustering

One of the principle objectives in occupational safety analysis is to identify the key factors that affect the severity of an incident. To identify risk groups of occupational incidents and the factors associated with them, statistical analysis of workers’ compensation claims data is performed using latent class clustering, for the segmentation of 1031 severe occupational incidents in agribusiness industries in the Midwest region of the United States between 2008–2016. In this study, severe incidents are those with workers’ compensation costs equal to or greater than $100,000 (USD). Based on the latent class clustering results, three risk groups are identified with injury nature as the most statistically distinctive classifier. The highest cost injuries include strain, tear, fracture, contusion, amputation, laceration, burn, concussion, and crushing. The most prevalent and statistically significant injury type is permanent partial disability. The study introduces a novel application of latent class clustering in the segmentation of high severity occupational incidents. The analytical approach and results of this study will aid safety practitioners in identifying occupational risk groups and analyzing injury patterns, and inform safety intervention plans to avoid the occurrence of similar incidents in agribusiness industries.

Big Data in occupational medicine: the convergence of -omics sciences, participatory research and e-health

Background:

New occupational hazards and risks are emerging in our progressively globalized society, in which ageing, migration, wild urbanization and rapid economic growth have led to unprecedented biological, chemical and physical exposures, linked to novel technologies, products and duty cycles. A focus shift from worker health to worker/citizen and community health is crucial. One of the major revolutions of the last decades is the computerization and digitization of the work process, the so-called “work 4.0”, and of the workplace.

Objectives:

To explore the roles and implications of Big Data in the new occupational medicine settings.

Methods:

Comprehensive literature search.

Results:

Big Data are characterized by volume, variety, veracity, velocity, and value. They come both from wet-lab techniques (“molecular Big Data”) and computational infrastructures, including databases, sensors and smart devices (“computational Big Data” and “digital Big Data”).

Conclusions:

In the light of novel hazards and thanks to new analytical approaches, molecular and digital underpinnings become extremely important in occupational medicine. Computational and digital tools can enable us to uncover new relationships between exposures and work-related diseases; to monitor the public reaction to novel risk factors associated to occupational diseases; to identify exposure-related changes in disease natural history; and to evaluate preventive workplace practices and legislative measures adopted for workplace health and safety.

International Occupational Health and Safety Management-Systems Standards as a Frame for the Sustainability: Mapping the Territory

A significant part of literature has shown that the adoption of Sustainability and Health-Safety management systems from organizations bears some substantial benefits since such systems (i) create a suitable frame for the sustainable development, implementation and review of the plans and/or processes, necessary to manage occupational health-safety (OHS) in their workplaces and (ii) imply innovative thinking and practices in fields of economics, policy-making, legislation, health and education. To this context, the paper targets at analysing current sustainability and OHSMSs in order to make these issues more comprehend, clear and functional for scholars and practitioners. Therefore, a literature survey has been conducted to map the territory by focusing on two interrelated tasks. The first one includes the presentation of the main International Management Systems (IMS) with focus on Sustainability and OHS (S_OHSMS) topics and the second task depicts a statistical analysis of the literature-review findings (for the years 2006–2017). In particular, the main purposes of the literature research were: (i) the description of key points of OHSMS and sustainability standards, (ii) the comparative analysis of their characteristics, taking into account several settled evaluation-criteria and (iii) the statistical analysis of the survey’s findings, while our study’s primary aim is the reinforcement of OHMSs’ application in any organization. The results evince, that the field of industry (with 28%) and also of the constructions (with 16%), concentrate the highest percentage of OHSMS use. In general, there were only few publications including OHSMSs (referred to various occupational fields) available in the scientific literature (during 2006–2017) but on the other hand, there was a gradually increasing scientific interest for these standards (especially during 2009–2012).

Analysing factors related to slipping, stumbling, and falling accidents at work: Application of data mining methods to Finnish occupational accidents and diseases statistics database

The utilisation of data mining methods has become common in many fields. In occupational accident analysis, however, these methods are still rarely exploited. This study applies methods of data mining (decision tree and association rules) to the Finnish national occupational accidents and diseases statistics database to analyse factors related to slipping, stumbling, and falling (SSF) accidents at work from 2006 to 2007. SSF accidents at work constitute a large proportion (22%) of all accidents at work in Finland. In addition, they are more likely to result in longer periods of incapacity for work than other workplace accidents. The most important factor influencing whether or not an accident at work is related to SSF is the specific physical activity of movement. In addition, the risk of SSF accidents at work seems to depend on the occupation and the age of the worker. The results were in line with previous research. Hence the application of data mining methods was considered successful. The results did not reveal anything unexpected though. Nevertheless, because of the capability to illustrate a large dataset and relationships between variables easily, data mining methods were seen as a useful supplementary method in analysing occupational accident data.