Framework for Multiobjective Optimization of Physical Highway Assets Investments

Abstract

Optimization-based tools have been included in many engineering management systems for individual infrastructure asset classes such as pavement management systems (PMS) and bridge management systems (BMS). These tools typically include single-objective optimization analyses. However, real-world decisions concerning asset preservation and renewal often involve more than one objective reflecting the various goals of the agency and need to evaluate possible alternatives according to multiple criteria. Traditional single-objective optimization approaches for handling such situations optimize a selected most important objective while either neglecting the less important competing objectives or imposing them as known constraints in the optimization formulation. This approach often does not provide truly optimal solutions. Multiobjective optimization formulations have clear theoretical advantages but increase the complexity of the mathematical formulation. This paper presents a review of the application of multiobjective optimization techniques in various working levels of highway asset management. Some promising techniques for the different infrastructure management functions are identified, and relevant characteristics are summarized and compared. Based on the applications reviewed, it can be concluded that multiobjective optimization could be effective for supporting many infrastructure management business processes. The review also suggests that a synergistic integration of complementary techniques may help develop practical and efficient decision-supporting tools that take advantage of the benefits and avoid potential drawbacks of the individual techniques.