A hybrid FAHP–Bayesian network framework for human reliability assessment in WSN based healthcare monitoring system

Abstract Wireless Sensor Networks (WSNs) are increasingly deployed in healthcare systems for continuous monitoring of critical patients. Despite technological advancements, human interaction during sensor installation, monitoring, alarm interpretation, and response introduces the risk of human error. Assessment of human reliability in such systems remains challenging due to uncertainty and limited operational data. This study proposes an integrated Fuzzy Analytic Hierarchy Process (FAHP) and Bayesian Network (BN) framework to estimate human error probability in WSN based healthcare monitoring systems. Human reliability factors and sub-factors are identified through past literature review. Expert judgments are expressed using linguistic variables and modelled using triangular fuzzy number. FAHP based on Chang’s extent analysis method is applied to determine weight of subfactors, and the resulting fuzzy failure probabilities are used as inputs to a BN to estimate overall human error probability. The proposed framework is demonstrated through a WSN based patient monitoring case study. The estimated overall human error probability is 0.1346 and human reliability value is 0.8654, indicating that human-related risks remain significant despite the presence of advanced monitoring technologies. The importance analysis shows that workload and stress, training and experience, and organizational and management factors are the most influential factors. Shift timing, confidence in interpreting automated sensor recommendations, frequency of refresher training, and long working hours are the most significant sub-factors that influence human error at the sub-factor level. The findings indicate that to enhance human reliability in healthcare WSN system, special consideration should be paid to the operational and organizational conditions. The proposed FAHP–BN framework provides a systematic and quantitative tool to support risk informed decision making and enhance patient safety in technology assisted healthcare environments.