Residual Life Prediction in the Presence of Human Error Using Machine Learning

Abstract

Accurate maintenance decision making is important for military equipment. Extremely demanding situations like limited time availability for maintenance during the war escalate the possibility of human errors in the maintenance of such equipment. Human errors in maintenance negatively impact the life of the systems. Human Reliability Analysis (HRA) methodologies have evolved to systematically quantify the human error in terms of Human Error Probability (HEP). However, the exact effect of the human error on the life of the component is unknown yet. In the presence of the diverse operating profiles for military equipment, estimating such effects becomes a complex and mathematically challenging problem to be handled by the conventional statistical techniques. This paper presents a machine learning approach to estimate the residual life of a component by incorporating the effect of human error in maintenance. Based on the nature of the maintenance data, a decision tree based boosted ensemble machine learning model is developed which predicts the Remaining Useful Life (RUL) of the component while considering error induced by maintenance personnel during its maintenance. The developed model is illustrated in the decision-making of replacement of a component in a mission critical military system in pre-mission maintenance break.