8th International Conference on Managing Pavement Assets REMOVING THE TYRANNY OF AVERAGES – MODELLING THE

The expectations towards asset management systems (AMS) often exceed what most AMS can or do deliver. The disappointment of stakeholders with an implemented AMS frequently leads to the abandonment of the otherwise sound principles of asset management. A review of the critical components of transport asset management systems highlighted several issues of which the treatment of uncertainties and the treatment selection method were explored in detail. Though performance forecasts based on average values may be right on the average, they are most likely wrong when looking at smaller segment levels in specific cases – i.e. most of the time. The key performance indicator for the Sydney Harbour Bridge maintenance requires that only a very small proportion of the paint is in need of immediate repair. To meet the specified maintenance objectives, a model is required that can contemplate both the extent and severity of the paint condition of each element of the Bridge. The paper discusses and describes the development of a suitable model based on Markov chains. The model has a broader application well beyond the Sydney Harbour Bridge as it is suitable for modelling condition parameters defined in terms of discrete condition states. The model relieves the users from the tyranny of the commonly used statistical parameters describing central tendencies, such as the average, thus making characterisation of an asset element more representative and thus accurate. As multiple conditions on the same element are possible, the accuracy of the budget and condition forecast also increases