Proper utilization of construction resources is critical to the success of highway projects, especially in post-disaster reconstruction situations. The lack of adequate construction resources places a great burden on decision makers to make a prudent use of these scarce resources in an efficient and effective manner. This includes deployment of limited resources to competing highway projects in such a way that minimizes the impact of construction works on network service disruption and construction costs. The literature is rich of research studies that addressed utilization of construction resources and they focused on two types of optimization problems: (1) single-objective optimization, and (2) multi-objective optimization.
2.3.1 Single-Objective Optimization
Many research studies focused on planning and optimizing the utilization of resources in construction projects with the objectives of either: (1) minimizing fluctuations in resource
requirements (resource leveling); or (2) resolving conflicts between activities/projects competing for the same resources (resource allocation).
First, a number of research studies tried to minimize the fluctuations in resource requirements and the negative impact these fluctuations have on construction productivity and cost. These studies used different optimization tools including: (1) heuristic methods (Ahuja 1976; Akpan 2000; Burgess and Killebrew 1962; Harris 1978); (2) linear programming (Easa 1989; Mattila and Abraham 1998); (3) integer programming (Son and Mattila 2004); (4) dynamic programming (Bandelloni et al. 1994); (5) simulated annealing (Son and Skibniewski 1999); (6) mathematical method (Senouci and Adeli 2001); and (7) genetic algorithms (Chan et al. 1996; Chua et al. 1996; Hegazy 1999; Leu and Yang 1999; Senouci and Eldin 2004).
Second, different research and optimization methods were utilized in an effort to allocate limited resources among activities/projects competing for the same type of resource, such as: heuristics, genetic algorithms (GA), dynamic programming, and particle swarm. Heuristic methods were used in several research studies to resolve conflicts between competing activities of a single project, especially in highway projects (Ahuja 1976; Bell and Han 1991; Boctor 1990; Sampson and Weiss 1993; El-Rayes and Moselhi 1998). Similarly, GAs are extensively used in the literature to: (1) optimize resource allocation with the single objective of minimizing project durations (Chan et al. 1996; and Chua et al. 1996); (2) solve large-scale resource allocation problems (Kim and Ellis 2008); (3) suggest modifications to genetic operators to better suit resource
optimization methods such as simulated annealing (Lee and Y. Kim 1996). Also, El- Rayes (2001) used a dynamic programming approach to develop a resource utilization optimization model for highway that utilize A+B bidding method with the objective of minimizing the total bid cost. Finally, Zhang et al. (2006) used particle swarm optimization to solve the same problem of minimizing project durations under resource constraints.
Despite the significant contributions of the above research studies, the resource utilization metrics and models developed are inadequate to deal with sharing limited resources among competing highway construction resource, especially in post-disaster situations. This is mainly due to the following characteristics of highway post-disaster reconstruction projects: (1) unusual large scope of work; (2) similarity of reconstruction resources and therefore high demand for specific types of resources; and (2) spatial dispersion of reconstruction projects over a large geographical area.
2.3.2 Multi-Objective Optimization
Genetic algorithms (GAs) have been extensively used in multi-objective optimization of resource utilization in construction, especially in highway projects. The planning objectives of these optimization problems include: (1) minimizing construction time and cost; (2) minimizing construction time and cost while maximizing quality; and (3) minimizing construction time and maximizing crew work continuity.
First, several research studies used GAs to perform construction time-cost trade-off analyses. For example, Feng et al. (1997) developed a GA-based spreadsheet for
analyzing time and cost of construction projects. Similarly, Marzouk and Moselhi (2004) used GAs with discrete event simulation and object-oriented programming to optimize earthmoving operations with the objective of simultaneously minimizing project cost and duration. Zheng et al. (2004) developed a GA-based multi-objective model for solving the time-cost trade-off problem, which uses a fitness function that factors in the values of time and cost of each chromosome using weights that adjust at every generation. In a following paper, Zheng and Ng (2005) integrated risk and uncertainty to the previous model to develop a stochastic approach to multi-objective optimization of time and cost in construction projects.
In addition to the use of GAs, ant colony optimization has also been used to solve time- cost tradeoff problems. Xiong and Kuang (2008) combined ant colony optimization with the modified adaptive weight approach proposed by Zheng et al. (2004) in order to generate optimal tradeoffs between project time and cost. Using the exact problems analyzed by Feng et al. (1997) and Zheng et al. (2004), ant colony optimization provided comparable if not better results that those generated by GAs (Xiong and Kuang 2008)
Second, El-Rayes and Kandil (2005) added a new dimension to the traditional time-cost tradeoff analysis in construction projects by trying to maximize construction quality. In this study, a multi-objective GA-based optimization model was developed to identify the optimal combination(s) of construction method, crew formation, and crew overtime policy that minimizes construction duration and cost while maximizing quality in highway
facilitate analyzing large-scale projects, parallel computing was utilized to reduce the computational time requirements for the GA-based time-cost-quality tradeoff analysis (Kandil and El-Rayes 2006a; Kandil and El-Rayes 2006b).
Finally, Hyari and El-Rayes (2006) developed a multi-objective optimization model to plan and schedule construction repetitive projects. This model is aimed at identifying the combination(s) of crew formation and crew interruption vectors that provide the optimal tradeoff between minimizing project duration and maximizing crew work continuity, simultaneously.
2.3.3 Limitation of Existing Research
Despite the significant contributions and practical features of the aforementioned research studies, further research is needed to cover the following needs in relation to allocating limited resources to competing post-disaster reconstruction projects:
- allocating multiple types of resource among competing reconstruction projects; - taking into account that reconstruction resources are available from different sources
(e.g. contractors) and at different times;
- studying the impact of project prioritization on reconstruction duration and cost; - identifying a practical methodology to assign highway post-disaster reconstruction
projects to qualified interested contractors; and
- considering the impact of working for extended hours and/or multiple shift on
productivity and therefore on highway construction duration and cost in post-disaster situations.