Occupational Injury Risk Mitigation: Machine Learning Approach and Feature Optimization for Smart Workplace Surveillance

Abstract

Forecasting the severity of occupational injuries shall be all industries’ top priority. The use of machine learning is theoretically valuable to assist the predictive analysis, thus, this study attempts to propose a feature-optimized predictive model for anticipating occupational injury severity. A public database of 66,405 occupational injury records from OSHA is analyzed using five sets of machine learning models: Support Vector Machine, K-Nearest Neighbors, Naïve Bayes, Decision Tree, and Random Forest. For model comparison, Random Forest outperformed other models with higher accuracy and F1-score. Therefore, it highlighted the potential of ensemble learning as a more accurate prediction model in the field of occupational injury. In constructing the model, this study also proposed the feature optimization technique that revealed the three most important features; ‘nature of injury’, ‘type of event’, and ‘affected body part’ in developing model. The accuracy of the Random Forest model was improved by 0.5% or 0.895 and 0.954 for the prediction of hospitalization and amputation, respectively by redeveloping and optimizing the model with hyperparameter tuning. The feature optimization is essential in providing insight knowledge to the Safety and Health Practitioners for future injury corrective and preventive strategies. This study has shown promising potential for smart workplace surveillance.

Systematic review of predictive maintenance and digital twin technologies challenges, opportunities, and best practices

“…Unsupervised learning is typically applied to clustering issues and looks for hidden patterns or intrinsic structures in the input data ( Abd-Elrazek et al, 2021 ). ML techniques such as naive Bayes, random forest, support vector machine (SVM), decision tree, and regression models are the most widely used AI approaches ( Khairuddin et al, 2022 , 2023 ; Neo et al, 2022 ), artificial neural network (ANN), and decision trees ( Hosamo et al, 2022 ) and ML is also used for Autoregressive Integrated Moving Average (ARIMA) ( Yang et al, 2021 ). From the review, 19 studies use ML approaches for PdM using DT ( Altun & Tavli, 2019 ; Avornu et al, 2022 ; Das et al, 2021 ; Harries et al, 2023 ; Heim et al, 2020 ; Hosamo et al, 2022 , 2023 ; Hu et al, 2023a ; Luo et al, 2020 ; Mourtzis, Tsoubou & Angelopoulos, 2023 ; Mubarak et al, 2022 ; Panagou et al, 2022a , 2022b ; Rajesh et al, 2019 ; Rossini et al, 2020 ; Siddiqui, Kahandawa & Hewawasam, 2023 ; Singh et al, 2023 ; Wang et al, 2022 ; Werner, Zimmermann & Lentes, 2019 ).…”

A Review of Data Mining Strategies by Data Type, with a Focus on Construction Processes and Health and Safety Management

“…In Occupational accident field, 16 out of 61 papers dealt with data on accidents and injuries at work from 2014 to 2023. In the class "occupational accidents," the 89 out of 202 submissions covered the following topics: reporting of accidents [3,26,27,29,32,35,43,45,49,[59][60][61][62]75,76] and days away from work. On this topic, Yelda et al analysed textual narratives to predict injury outcome and days off work in a mining operation.…”

“…The above criteria allowed us to assign a "label" to each component. (32), knn (49), svm (81) and rf (69), to the right of the graph characterised by a strongly positive coordinate on the axis, to individuals such as MCDA C (58), characterised by a strongly negative coordinate on the axis (to the left of the graph). Dimension 2 opposes individuals such as lstm (54), word2vec (88), nlp (63) and BIM ( 16), who at the top of the graph, and characterised by a low positive co-ordinate on the axis, with individuals such as ann (8), adaboost (3), who have low negative coordinate on the axis and are located at the bottom of the graph.…”

Prediction of Safety Performance by Using Machine Learning Algorithms: Evidence from Indian Construction Project Sites

“…Finally, the RF classifier with 95% accuracy, 100% precision, 95% recall, 97.5% f-measure and auc score 96.0% outperformed from the results of other algorithms. RF classifier outperformed the Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Naïve Bayes (NB) and DT in predicting the occupational risk injury severity with higher accuracy and f1-score (Khairuddin et al, 2022).…”