Assessment of HEMM Operators’ Risk Exposure due to Whole-Body Vibration in Underground Metalliferous Mines Using Machine Learning Techniques

Abstract

Whole-body vibration (WBV) is a substantial occupational health and safety hazard to heavy earth-moving machinery (HEMM) operators. There is a need to appraise the effect of WBV jeopardize and the factors influencing the WBV risk exposure on the HEMM operators. Seven machine learning (ML) models were tested on 81 data samples collected from seven underground metalliferous mines. The study considered nine factors which have substantial role behind the intensity of the WBV risk exposure of HEMM operators. RReleifF algorithm was used for dimensionality reduction and ranking the features. Compared to other ML techniques, ANN model was determined to be the most effective approach. The nine considered features were reduced to five features using RReleifF algorithm. The ranking of the five features selected was in order of awkward posture, the machine age, haul road condition, speed, and seat thickness based on their weights. Finally, a predictive equation was developed using the aforementioned five features. This study will help the seven underground mines authority to evaluate the WBV risk exposure effortlessly without the usage of scientific instrument and also helps in adopting immediate control measures to mitigate WBV risk exposure of HEMM operators.