Military Facility Cost Estimation System Using Case-Based Reasoning in Korea

Numerous cost estimations are made repetitively in the initial stages of construction projects in response to ongoing scope changes and often need to be recalculated frequently. In practice, the square foot method, considered an effective method for time-saving, is widely used. However, this method requires a great amount of effort to calculate a unit price and does not consider the uniqueness of each case. Thus, the use of the square foot method could bring about unwanted consequences. For example, in the case of military projects in Korea, significant differences have been reported between estimations made using this method and the actual costs. In an effort to deal with this challenging issue, this research develops a military facility cost estimation (MilFaCE) system, based on case-based reasoning (CBR), using case data from 422 construction projects at 16 military facilities. Based on system validation experiments involving 10 military officers (engineers), the effectiveness of the system in terms of estimation accuracy and user-friendliness is confirmed. Consequently, this research can be a CBR application example of construction cost estimation and a basis for further research into the development of cost estimate systems.     

Hybrid Models of Neural Networks and Genetic Algorithms for Predicting Preliminary Cost Estimates

This technical note applies hybrid models of neural networks (NN) and genetic algorithms (GA) to cost estimation of residential buildings to predict preliminary cost estimates. Data used in the study are for residential buildings constructed from 1997 to 2000 in Seoul, Korea. These are used in training each model and evaluating its performance. The models applied were Model I, which determines each parameter of a back-propagation network by a trial-and-error process; Model II, which determines each parameter of a back-propagation network by GAs; and Model III, which trains weights of NNs using genetic algorithms. The research revealed that optimizing each parameter of back-propagation networks using GAs is most effective in estimating the preliminary costs of residential buildings. Therefore, GAs may help estimators overcome the problem of the lack of adequate rules for determining the parameters of NNs.     

Applying Pareto Ranking and Niche Formation to Genetic Algorithm-Based Multiobjective Time–Cost Optimization

Time–cost optimization (TCO) is one of the greatest challenges in construction project planning and control, since the optimization of either time or cost, would usually be at the expense of the other. Although the TCO problem has been extensively examined, many research studies only focused on minimizing the total cost for an early completion. This does not necessarily convey any reward to the contractor. However, with the increasing popularity of alternative project delivery systems, clients and contractors are more concerned about the combined benefits and opportunities of early completion as well as cost savings. In this paper, a genetic algorithms (GAs)-driven multiobjective model for TCO is proposed. The model integrates the adaptive weight to balance the priority of each objective according to the performance of the previous “generation.” In addition, the model incorporates Pareto ranking as a selection criterion and the niche formation techniques to improve popularity diversity. Based on the proposed framework, a prototype system has been developed in Microsoft Project for testing with a medium-sized project. The results indicate that greater robustness can be attained by the introduction of adaptive weight approach, Pareto ranking, and niche formation to the GA-based multiobjective TCO model.     

Minimizing Cutting Wastes of Reinforcement Steel Bars Using Genetic Algorithms and Integer Programming Models

Materials that are in the form of one-dimensional stocks such as steel rebars, structural steel sections, and dimensional lumber generate a major fraction of the generated construction waste. Cutting one-dimensional stocks to suit the construction project requirements result in trim or cutting losses, which is the major cause of the one-dimensional construction waste. The optimization problem of minimizing the trim losses is known as the cutting stock problem (CSP). In this paper, three approaches for solving the one-dimensional cutting stock problem are presented. A genetic algorithm (GA) model, a linear programming (LP) model, and an integer programming (IP) model were developed to solve the one-dimensional CSP. Three real life case studies from a steel workshop have been studied. The generated cutting schedules using the GA, LP, and IP approaches are presented and compared to the actual workshop’s cutting schedules. The comparison shows a high potential of savings that could be achieved using such techniques. Additionally, a user friendly Visual Basic computer program that utilizes genetic algorithms for solving the one-dimensional CSP is presented.     

Estimation of Cost Contingency for Air Force Construction Projects

Risk and associated cost overruns are critical problems for construction projects, yet the most common practice for dealing with them is the assignment of an arbitrary flat percentage of the construction budget as a contingency fund. Therefore, our goal was to identify significant variables that may influence, or serve as indicators of, potential cost overruns. We analyzed data from 203 Air Force construction projects over a full range of project types and scopes using multiple linear regression to develop a model to predict the amount of required contingency funds. The proposed model uses only data that would be available prior to the award of a construction contract. The variables in the model were categorized as project characteristics, design performance metrics, and contract award process influences. Based on the performance metric used, the model captures 44% of actual cost overruns versus the 20% captured by the current practice. Furthermore, application of the model reduces the average contingency budgeting error from 11.2 to only 0.3%.     

Finance-Based Scheduling: Tool to Maximize Project Profit Using Improved Genetic Algorithms

Contractor’s ability to procure cash to carry out construction operations represents a crucial factor to run profitable business. Bank overdrafts have always been the major source to finance construction projects. However, it is not uncommon that bankers set a limit on the credit allocated to an established overdraft. Bankers’ interest rates and consequently contractors’ financing costs are basically determined based on the allocated credit limits. Furthermore, project indirect costs are directly proportional to the project duration which is affected by the allocated credit limit. Thus, the credit limit affects project financing costs and indirect costs which in turn affect project profit. However, finance-based scheduling produces financially executable schedules at specified credit limits while maintaining the demand of time minimization. Thus, finance-based scheduling provides a tool to control the credit requirements. This control enables contractors to negotiate lower interest rates which reduce financing costs. Thus, finance-based scheduling enables contractors to reduce project indirect costs and financing costs. This paper utilizes genetic algorithm’s technique to devise finance-based schedules that maximize project profit through minimizing financing costs and indirect costs.     

Improving Cost Estimates of Construction Projects Using Phased Cost Factors

Central to cost-based competition is the capability to accurately predict the cost of delivering a project. Most literature on cost estimation focuses on specific estimation methods as generic techniques and little attention has been paid to the unique requirements at each project stage. This note attempts to identify the critical factors for effective estimation at various stages of typical construction projects. Drawing from organization control theory and cost estimating literature, this note develops a theoretical framework that identifies the critical factors for effective cost estimation during each project phase of a conventional construction project. The underlying logic is that as a cost estimating effort progresses, both task programmability and output measurability improve. As a result, control effort will shift from input-oriented control to a combination of output and behavior control.     

Predicting Construction Cost Using Multiple Regression Techniques

This paper describes the development of linear regression models to predict the construction cost of buildings, based on 286 sets of data collected in the United Kingdom. Raw cost is rejected as a suitable dependent variable and models are developed for cost/m2, log of cost, and log of cost/m2. Both forward and backward stepwise analyses were performed, giving a total of six models. Forty-one potential independent variables were identified. Five variables appeared in each of the six models: gross internal floor area (GIFA), function, duration, mechanical installations, and piling, suggesting that they are the key linear cost drivers in the data. The best regression model is the log of cost backward model which gives an R2 of 0.661 and a mean absolute percentage error (MAPE) of 19.3%; these results compare favorably with past research which has shown that traditional methods of cost estimation have values of MAPE typically in the order of 25%.     

Resource Optimization Using Combined Simulation and Genetic Algorithms

This paper presents a new approach for resource optimization by combining a flow-chart based simulation tool with a powerful genetic optimization procedure. The proposed approach determines the least costly, and most productive, amount of resources that achieve the highest benefit/cost ratio in individual construction operations. To further incorporate resource optimization into construction planning, various genetic algorithms (GA)-optimized simulation models are integrated with commonly used project management software. Accordingly, these models are activated from within the scheduling software to optimize the plan. The result is a hierarchical work-breakdown-structure tied to GA-optimized simulation models. Various optimization experiments with a prototype system on two case studies revealed its ability to optimize resources within the real-life constraints set in the simulation models. The prototype is easy to use and can be used on large size projects. Based on this research, computer simulation and genetic algorithms can be an effective combination with great potential for improving productivity and saving construction time and cost.     

Bridge Model Updating Using Response Surface Method and Genetic Algorithm

A finite-element (FE) model of a structure is a highly idealized engineering model that may or may not truly reflect the physical structure. The purpose of model updating is to modify the FE model of a structure in order to obtain better agreement between the numerical and field-measured structure responses. In this paper, a new practical and user-friendly FE model updating method is presented. The new method utilizes the response surface method for the best experimental design of the parameters to be updated based on which numerical analysis can be performed in order to obtain explicit relationships between the structural responses and parameters from the simulation results. The parameters are then be updated using the genetic algorithm (GA) by minimizing an objective function. A numerical example of a simply supported beam has been used to demonstrate the concept. This method has also been applied to the model updating of an existing bridge. Results show that this method works well and achieves reasonable physical explanations for the updated parameters.     

Optimal Reservoir Operation for Irrigation of Multiple Crops Using Genetic Algorithms

This paper presents a genetic algorithm (GA) model for obtaining an optimal operating policy and optimal crop water allocations from an irrigation reservoir. The objective is to maximize the sum of the relative yields from all crops in the irrigated area. The model takes into account reservoir inflow, rainfall on the irrigated area, intraseasonal competition for water among multiple crops, the soil moisture dynamics in each cropped area, the heterogeneous nature of soils, and crop response to the level of irrigation applied. The model is applied to the Malaprabha single-purpose irrigation reservoir in Karnataka State, India. The optimal operating policy obtained using the GA is similar to that obtained by linear programming. This model can be used for optimal utilization of the available water resources of any reservoir system to obtain maximum benefits.     

Object-Oriented Optimization Approach Using Genetic Algorithms for Lattice Towers

A new approach is presented for the optimization of steel lattice towers by combining genetic algorithms and an object-oriented approach. The purpose of this approach is to eliminate the difficulties in the handling of large size problems such as lattice towers. Improved search and rapid convergence are obtained by considering the lattice tower as a set of small objects and combining these objects into a system. This is possible with serial cantilever structures such as lattice towers. A tower consists of panel objects, which can be classified as separate objects, as they possess an independent property as well as inherent properties. This can considerably reduce the design space of the problem and enhance the result. An optimization approach for the steel lattice tower problem using objects and genetic algorithms is presented here. The paper also describes the algorithm with practical design considerations used for this approach. To demonstrate the approach, a typical tower configuration with practical constraints has been considered for discrete optimization with the new approach and compared with the results of a normal approach in which the full tower is considered.     

Optimization of Pile Groups Using Hybrid Genetic Algorithms

This paper presents an automated optimal design method using a hybrid genetic algorithm for pile group foundation design. The design process is a sizing and topology optimization for pile foundations. The objective is to minimize the material volume of the foundation taking the configuration, number, and cross-sectional dimensions of the piles as well as the thickness of the pile cap as design variables. A local search operator by the fully stressed design (FSD) approach is incorporated into a genetic algorithm (GA) to tackle two major shortcomings of a GA, namely, large computation effort in searching the optimum design and poor local search capability. The effectiveness and capability of the proposed algorithm are first illustrated by a five by five pile group subjected to different loading conditions. The proposed optimization algorithm is then applied to a large-scale foundation project to demonstrate the practicality of the algorithm. The proposed hybrid genetic algorithm successfully minimizes the volume of material consumption and the result matches the engineering expectation. The FSD operator has great improvement on both design quality and convergence rate. Challenges encountered in the application of optimization techniques to design of pile groups consisting of hundreds of piles are discussed.     

Comparison of Using Mixed-Integer Programming and Genetic Algorithms for Construction Site Facility Layout Planning

The use of modular construction has gained wide acceptance in the industry. For a specific construction facility layout problem such as site precast standardized modular units, it requires the establishment of an on-site precast yard. Arranging the precast facilities within a construction site presents real challenge to site management. This complex task is further augmented with the involvement of several resources and different transport costs. A genetic algorithm (GA) model was developed for the search of a near-optimal layout solution. Another approach using mixed-integer programming (MIP) has been developed to generate optimal facility layout. These two approaches are applied to solve with an example in this paper to demonstrate that the solution quality of MIP outperforms that of GA. Further, another scenario with additional location constraints can also be solved readily by MIP, which, however, if modeled by GA, the solution process would be complicated. The study has highlighted that MIP can perform better than GA in site facility layout problems in which the site facilities and locations can be represented by a set of integer variables.     

Using Decision Trees for Determining Attribute Weights in a Case-Based Model of Early Cost Prediction

This paper compares the performance of three different decision-tree-based methods of assigning attribute weights to be used in a case-based reasoning (CBR) prediction model. The generation of the attribute weights is performed by considering the presence, absence, and the positions of the attributes in the decision tree. This process and the development of the CBR simulation model are described in the paper. The model was tested by using data pertaining to the early design parameters and unit cost of the structural system of residential building projects. The CBR results indicate that the attribute weights generated by taking into account the information gain of all the attributes performed better than the attribute weights generated by considering only the appearance of attributes in the tree. The study is of benefit primarily to researchers, as it compares the impact of attribute weights generated by three different methods and, hence, highlights the fact that the prediction rate of models such as CBR largely depends on the data associated with the parameters used in the model.     

Construction Planning Method Using Case-Based Reasoning (CONPLA-CBR)

During the construction process, there occur many unexpected events that hinder timely completion of a project. One plausible solution in formulating a robust plan against such uncertainties is to provide the proper construction duration by utilizing as-built schedules in which past events are stored for similar future projects. Many schedulers thus develop schedules based on similar past schedules, taking into consideration the dynamic construction environment. As a result, construction schedulers normally refer to similar past schedules for their current projects. Few studies on the reuse of past schedules using case-based reasoning (CBR) have been conducted, and those that are available are limited to specific areas of construction such as apartment-building construction and boiler manufacturing. This research has an emphasis on developing a CBR-based general planning tool with higher applicability, which consists of generic attributes with the capability to be customized to the given project. To address this issue, construction planning by CBR (CONPLA-CBR) is presented as a generic planning tool for various types of construction projects. CONPLA-CBR, with the dynamic case approach and construction schedule data mart, also developed in this research, helps schedulers to utilize past schedules. CONPLA-CBR was not only verified to be of practical utility by experts, but also, because it uses past cases to which the successor relationships pertain, it does not require users to input all relationships. Whereas the proposed CONPLA-CBR generates master schedules at the preconstruction stage, its concept can also be applied to the construction stages to generate more detailed, for example, weekly or monthly, schedules. Thereby, CONPLA-CBR would enhance construction performance through the increased application of CBR in construction.     

Genetic Algorithms in Competitive Environments

Competition is introduced among the populations of a number of genetic algorithms (GAs) in solving optimization problems. The aim is to adapt the parameters of the GAs, by altering the resources of the system, so as to achieve better solutions. The evolution of the different populations, having different sets of parameters, is controlled at the level of metapopulation, i.e., the union of populations, on the basis of statistics and trends of the evolution of every population. An overall fitness measure is introduced that incorporates a diversity measure and the required resources to rank the populations. The fuzzy outcome of the conflict among the populations guides the evolution of the different GAs toward better solutions in the statistical sense. The proposed scheme is applied to two different problems—a multimodal function with six global and several near-global optima, and a reliability based optimal design of a simple truss. Numerical results are presented, and the robustness and computational efficiency of the proposed scheme are discussed.     

CBR Revision Model for Improving Cost Prediction Accuracy in Multifamily Housing Projects

For any construction project to succeed, it is very important to accurately estimate the construction cost during the project’s initial stage. This is why there has been much interest lately in cost prediction models that use case-based reasoning (CBR). It has been pointed out, however, that existing CBR-based cost prediction models may yield inaccurate results even though they could survey optimal similar cases, if the number of cases in the case base is not enough. As opposed to the existing CBR-based construction cost prediction models, this study developed a CBR revision model that reflects the “revise” phase of the CBR cycle (retrieve, reuse, revise, and retain) based on nine multifamily housing projects executed recently by “A” Housing Corporation. To verify the developed model, a case study was performed using three case projects completed by “B” and “C” Housing Corporations. The result showed that the prediction error ratio after the Revise (I) phase decreased from 97.44 to 22.58%. This model can be effective when there are insufficient established cases in the case base.     

Trihalomethane Species Forecast Using Optimization Methods: Genetic Algorithms and Simulated Annealing

Chlorination is an effective method for disinfection of drinking water. Yet chlorine is a strong oxidizing agent and easily reacts with both organic and inorganic materials. Trihalomethanes (THMs), formed as a by-product of chlorination, are carcinogenic to humans. Models can be derived from linear and nonlinear multiregression analyses to predict the THM species concentration of empirical reaction kinetic equations. The main objective of this study is to predict the concentrations of THM species by minimizing the nonlinear function, representing the errors between the measured and calculated THM concentrations, using the genetic algorithm (GA) and simulated annealing (SA). Additionally, two modifications of SA are employed. The solutions obtained from GA and SA are compared with the measured values and those obtained from a generalized reduced gradient method (GRG2). The results indicate that the proposed heuristic methods are capable of optimizing the nonlinear problem. The predicted concentrations may provide useful information for controlling the chlorination dosage necessary to assure the safety of water drinking.     

Genetic Algorithms for Construction Site Layout in Project Planning

Construction site layout is concerned with the existence, positioning, and timing of the temporary facilities that are used to carry out a construction project. Typically these problems are very complicated to formulate and difficult to solve. They are, however, very important to virtually any construction project, since the site layout can significantly affect the cost of the project. This paper describes the general site layout problem from both a theoretical and a practical point of view. It proposes genetic algorithms as a possible solution technique and includes a theoretical example of positioning temporary facilities. This is extended to a practical problem in which the cost of movement is modeled realistically using an augmented genetic algorithm. Some preliminary conclusions are drawn for the application of genetic algorithms to construction site layout problems.