A resilience-based maintenance optimisation framework using multiple criteria and Knapsack methods

Ahmed Noaman Karar

doi:10.1016/j.ress.2023.109674

Return

A resilience-based maintenance optimisation framework using multiple criteria and Knapsack methods

2024-01-01

PRE

Ahmed Noaman Karar *

Ashraf Labib

Dylan Jones

DOI：10.1016/j.ress.2023.109674

Original

Original request for help

Save

Abstract

En 中文

Business fluctuations and pandemics such as COVID 19 have revealed the need for more resilient approaches and processes in the asset management domain. This research aims to design a resilience-based maintenance optimisation (RbMO) framework that absorbs the fluctuations in the operating context and sustains asset performance at an optimum maintenance cost and an acceptable level of risk. The paper proposes a framework that employs the analytical hierarchy process (AHP) to translate the different operating context parameters into risk aspects with relative weights that differ from one operating scenario to another. The Knapsack method then uses these relative weights to define the risk reduction of each maintenance task and pick the optimum ones within the allocated maintenance budget. Additionally, the approach introduces the nested criticality grid (NCG), which graphically demonstrates the inherent, Knapsack and residual risk profiles from the failure mode level up to the unit level enabling an informative decision-making process, where the asset owner can wisely distribute the maintenance budget or achieve efficient cost savings. Finally, the main advantage of this model lies in its comprehensive continuous improvement cycle of the proposed RbMO framework, which ties the different components of the proposed approach together and forms an integrated system that promotes a more resilient strategy for asset maintenance management, enabling asset owners to effectively manage business fluctuations and operating context changes.

Keywords:

Analytical hierarchy process

Knapsack method

Maintenance optimisation

Data-driven decision making