The U.S. Moving Ahead for Progress in the 21st Century Act (MAP-21) mandates the development of a risk-based transportation asset management plan and the use of a performance-based approach in transportation planning and programming. This paper introduces a systematic element-based multi-objective optimization (EB-MOO) methodology integrated into a goal-driven transportation asset management framework to improve bridge management and support state departments of transportation with their transition efforts to comply with these MAP-21 requirements. The methodology focuses on the bridge asset class and is structured around five modules: data processing, improvement, element-level optimization, bridge-level optimization, and network-level optimization modules. It relies on a leading-edge forecasting model, three separate screening processes (i.e., the element deficiency, alternative feasibility, and solution superiority screening processes) to overcome computer memory and processing time limitations, and a simulation arrangement to generate life-cycle alternatives (series of improvement actions). Additionally, the EB-MOO methodology consists of three levels of optimization assessment based on the Pareto optimality concept: element-level, bridge-level, and network-level (following either a top-down or bottom-up approach). A robust metaheuristic genetic algorithm handles the different multi-objective optimization problems. A prototyping tool was developed for the implementation of the methodology through several examples of unconstrained and constrained (by budget, performance, or both) scenarios. Results reveal the capability of the methodology to generate Pareto optimal or near-optimal solutions, predict performance, and determine funding requirements and short- and long-term intervention strategies detailed at the bridge-element level for planning and programming. The EB-MOO methodology can also be expanded to accommodate other asset classes or modes.

The paper briefly introduces an element-based multi-objective optimization (EB-MOO) methodology to support state departments of transportation with their decision-making process, asset management, and performance-based transportation planning and programming. The methodology focuses on the bridge asset class and consists of five modules: (i) data processing, (ii) improvement, (iii) element-level optimization (ELO), (iv) bridge-level optimization (BLO), and (v) network-level optimization (NLO) modules. These five modules jointly produce short- and long-term intervention strategies detailed at the bridge element level for planning and programming. The paper focuses on the BLO module, specifically: the basic framework of underlying processes and concepts, the optimization problem types and mathematical formulations, and the heuristic algorithm to solve the BLO problems. A prototyping tool is developed to implement these five modules of the EB-MOO methodology, test concepts, prove effectiveness, and demonstrate potential benefits. The paper also includes an illustrative example using the prototyping tool. The example consists of the BLO problems under different budget and/or performance scenarios. The implementation proves the module’s capability in producing a diverse set of Pareto optimal or near-optimal solutions, recommending set of element intervention actions and timings, predicting performance, and determining budget requirements for the entire program period. The BLO results associated with the recommended solutions serve as the fundamental inputs for the NLO module. Nevertheless, the BLO module can be used independently, providing a systematic process for the development of bridge improvement/preservation programs detailed at the element level.

The overall objective of this research is to support state departments of transportation with their decision-making processes and transitions to performance management and performance-based planning and programming mandated by the Moving Ahead for Progress in the 21st Century Act. Accomplishing this objective requires a systematic multiobjective optimization methodology. This research proposes such a methodology, referred to as an “element-based multiobjective optimization” (EB-MOO) methodology, which produces optimal or near-optimal sets of short- and long-term intervention strategies detailed at the bridge element level for planning and programming. The methodology currently focuses on the bridge asset class and consists of five modules: (1) data processing, (2) improvement, (3) element-level optimization (ELO), (4) bridge-level optimization (BLO), and (5) network-level optimization (NLO) modules. This paper details the ELO module, specifically: the basic framework of underlying processes and concepts, the alternative feasibility screening process, optimization problem types and mathematical formulations, and the heuristic algorithm used to solve the ELO problems. The paper also includes an illustrative example using a prototyping tool developed to implement EB-MOO methodology. The example presents several ELO problems under unconstrained scenarios. The implementation demonstrated the module’s capability in producing optimal or near-optimal ELO solutions, recommending element intervention actions, predicting performance, and determining funding requirements for the specified improvement type and program year. The broader EB-MOO methodology uses the ELO results as inputs for the BLO and NLO modules.