Andrew Colin

Optimizing Budget Allocations in Naval Configuration Management

This paper describes a case study illustration of a technique that can be used to optimize capital expenditure budget decisions for engineering assets under multivariate objectives and uncertain costs. We define measures of capability increase relative to mission types and measures for risk, return and pseudo-correlation on potential maintenance upgrades, allowing portfolio optimization techniques from finance to be applied to asset management decisions. The results provide a risk-reward ratio for investment allocation decisions that is optimal. The work is aimed at capability management for naval platforms, but the technique has broader applicability in the field of asset management and multi-criteria decision making.

Statistical Activity Cost Regression Analysis of a Scheduling Problem

Regression analysis of accounting costs on physical activity levels for a number of vessels is used to estimate optimum cost and capability of fleet schedules. Quarterly accounting and activity data is used, the regression analysis showing the existence of lags of 3–9 months between the occurrence of an engineering activity and the recording of the associated costs. When the lags are taken into account simple multivariable linear regression models of the fixed-variable cost type provide indicators of future costs based upon planned engineering activity levels. This information is employed in Monte Carlo experiments to determine minimum cost, maximum engineering capability, fleet maintenance and deployment schedules.