A consensus-based AHP for improved assessment of resilience engineering in maintenance organizations

Resilience engineering (RE) is a new approach to measuring and maintaining safety in complex systems. The focus of RE is not on errors rather than on understanding and supporting normal work and what goes right. Using analytical hierarchy process (AHP), the present study aims to devise a validated method for improved assessment of RE in maintenance organizations. A standardized questionnaire containing RE and Performance Shaping Factors (PSFs) for generic maintenance operators is designed to collect data from employees in 11 regional maintenance departments of a large public gas company. Holding regular discussion sessions with experts in the field, the AHP is then built up based on the consensus emerged from the discussions. To form the middle-level criteria of the analytical hierarchy, RE items are clustered into 6 new categories using a verified k-means clustering. Given the large number of RE items in some categories, a complete sensitivity analysis is then performed by data envelopment analysis (DEA) to identify the most important items as the final level criteria. The designed AHP is used to assess RE in the 11 regional maintenance departments. For verification and validation of the proposed AHP, the linear relationship between AHP-based RE assessment and PSFs assessment is tested for its significance. The results confirm the close relationship between RE and PSFs.