Application of Shackle's Model and System Optimization for Highway Investment Decision Making under Uncertainty

This paper introduces a new methodology for highway investment decision making under uncertainty that uses Shackle's model for uncertainty-based project benefit analysis and system optimization for project selection. Shackle's model overcomes limitations of the risk-based life-cycle cost analysis approach by using degree of surprise as a measure of uncertainty associated with possible outcomes of performance measures utilized for project benefit analysis instead of probability distribution. It also employs a priority index that jointly evaluates each possible outcome and degree of surprise pair for a performance measure to establish the standardized focus gain value and focus loss value from the expected outcome. The standardized focus gain-over-loss ratios corresponding to multiple performance measures are synthesized into an overall ratio that is regarded as the overall benefits of a highway project under uncertainty. This ratio is used as the basis of project selection. An optimization model is formulated as the multichoice multidimensional knapsack problem for project selection. A case study reveals significant differences in project selection results between the use of the proposed methodology and the existing risk-based approach.