Prioritisation in infrastructure asset maintenance

The aging infrastructure and rising demand in the irrigation industry as a result of population growth have increased maintenance works in recent years. The most efficient asset maintenance practice is proactive. However, limited budgets and an increase in aging infrastructure has made proactive asset maintenance challenging. Yet customers still expect quality service and contemporary challenges such as climate change and the competitive market just add to the existing pressure on asset owners.

The common risk-score based prioritisation (RSBP) method, which is used in practice to prioritise assets for maintenance is inaccurate and inefficient, due to the following reasons:

1. Existence of several qualitative and quantitative criteria within the maintenance prioritisation assessment. 2. Presence of uncertainty attached to subjective judgments by asset experts in a qualitative assessment. 3. Presence of conflicting cost and benefit criteria within the same assessment which complicates the calculations.

In this context, the present work has the primary objective of developing a novel, accurate, efficient and straightforward methodology for measuring criteria weights and ranking asset alternatives for maintenance. For this aim, multi-criteria decision-making (MCDM) methods are developed for optimisation of asset maintenance prioritisation. First, two compensatory MCDM methods - AHP (Analytical Hierarchy Process) and TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution), both in a fuzzy environment - are developed and for the first time applied to prioritise irrigation assets for maintenance. The fuzzy theory addresses the uncertainty of the subjective nature of the decision problem and therefore helps with optimisation of asset maintenance prioritisation. Fuzzy AHP, which is based on pairwise comparison matrices, measures the weights of criteria which then are utilised in fuzzy TOPSIS to rank the alternatives for maintenance. Three other compensatory MCDM methods - WSM (Weighted sum Model), WPM (Weighted Product Model) and COPRAS (COmplex PRoportional ASessment), also in a fuzzy environment - are then applied to the same problem to investigate further, verify and validate the prioritisation results.

In applying the newly developed MCDM method on the case study, seven evaluation criteria including service delivery, financial loss, safety, environment, credibility, compliance and asset condition rating (extracted from the case study risk framework) are evaluated by three asset management experts (two from asset management team and one from asset maintenance delivery team) from the same organisation to measure the criteria weights using fuzzy AHP method. The obtained weights by the newly developed method present more diverse distribution amongst the evaluation criteria compared to the weights measured by the existing prioritisation method in practice. For example, in the current evaluation practice, service delivery, credibility and environment has been assigned the same weight (or with very minimal difference) whereas the developed fuzzy AHP method's result seems more diverse including 23.6%, for service delivery, 9.95% for credibility and 8.46% for environment. Beyond fuzzy AHP method's capability to take uncertainty into account, fuzzy weights are graphically presented in chapter 5 to give readers' a visual presentation of footprint of uncertainty in determining the criteria weights, it is capable of identifying inconsistencies in evaluations to avoid 'garbage in garbage out' scenario in asset maintenance prioritisation. The newly developed fuzzy AHP-TOPSIS method is utilised to prioritise six irrigation channels for maintenance. For validating the prioritisation results of the developed fuzzy AHP-TOPSIS method, the same (six) irrigation channels are ranked by other compensatory MCDM methods as well as the asset management experts. The experts' ranking preferences are based on the physical condition, capacities (design capacities vary between 35 to 200 megalitre per day) and subjective judgment of their failure impacts in terms of considered evaluation criteria. The final ranking results by all MCDM methods have a high correlation with the preferences of asset management experts. A determining factor in this achievement is the capability of the developed fuzzy MCDM method in accommodating objective data (i.e. percentages of different tiers of asset condition ratings) in the assessment. As an example, channels no.1 and 4 which are identified/preferred as the first and last priority for maintenance by all MCDM methods as well as the asset management experts have respectively the worst and healthiest asset condition ratings which are objective values and therefore righteously can impact the result of the decision. The experimental practice demonstrated the excellent capability of the developed method in distinguishing the failure impacts of different alternatives in terms of decision criteria as almost all utilised MCDM method results had a high correlation (0.76 to 0.91 for all considered MCDM methods) with the experts' ranking preferences. This shows the importance of utilising more advanced methods like fuzzy MCDM to prioritise asset maintenance considering current budget limitations. Also, development of such method in academic studies can be utilised in smart asset maintenance prioritisation such as digital twin in asset management. In fact, the next step can be utilising fuzzy MCDM method to prioritise possible prescriptions (for treating assets) and present a fully automated and smart asset management method which is capable of sensing, identifying, analysing (based on live data) and prescribing solutions (in a subjective environment) to maintain assets.

The present work provides a new decision-making model for infrastructure maintenance prioritisation. The major outcomes of this effort are the ability to use qualitative and quantitative cost and benefit criteria in the same assessment, addressing the uncertainty in subjective environments by utilising fuzzy theory, and developing a user-friendly software based on the fuzzy AHP-TOPSIS method. This software automates the developed method¿s calculations and therefore increases the likelihood of the method's application in real-world case scenarios.