Feature-based Construction Cost Estimating

By Sheryl Staub-French

Cost estimates are created an average of seven times throughout the life of a project (Laitinen 1998)*.  Consequently, it is critical that estimating tools not only support the generation of cost estimates, but also assist estimators with the maintenance of cost estimates throughout the project life-cycle.  Current estimating software helps estimators take off quantities automatically and establish an electronic link between a component in a 3D-product model (specific to a CAD environment or in an IFC model) and a cost item in a cost estimating database to create a cost estimate.  Quantity-based relationships between product and cost information are helpful in assisting estimators with the creation of estimates.  However, they are insufficient in helping estimators to maintain the cost estimates as the design evolves and changes throughout the course of the project.  The reason is that in addition to material quantities, product features, such as openings and repetition, also affect construction costs.  Moreover, existing software tools do not explicitly capture estimators' rationale for how the component properties and product features affect the cost information.  Without computer support to store and use these meta-data about the relationships between product and cost information, estimators will need to determine when and how to adjust the cost information in the case of design changes so that the cost estimate and project scope are in balance.  Hence, I outline new estimating functionality and the corresponding formalisms and mechanisms needed to identify product features and properties and their effects on construction costs to assist estimators with the maintenance of cost estimates.

The practical motivation for this research is based on our experience working with a project team throughout design and construction of a pilot plant facility for Sequus Pharmaceuticals (Staub and Fischer 2000).  We will describe the drywall subcontractor's estimating process to determine the labor and material costs for all four walls of the room shown in Figure 1.  Specifically, we will illustrate what component properties and product features are important to the estimator, how they affect the construction cost information, and how well current tools help the estimator to maintain the cost estimate as the design changes (example product features and properties are annotated in Figure 1).
 
 

Figure 1: Current estimating process using tools that link 
design and construction costs 

Figure 1 shows an overview of the current estimating process using commercial estimating software that links with 3D CAD software.  The steps to create an estimate are as follows:
1) Create Cost Assembly: First, the estimator creates a cost assembly to aggregate all the estimating items that are needed for each wall in the room.  Each item in the assembly contains information about the material unit costs, labor and equipment resources, and the production rates. 
2) Apply Cost Assembly: After the assembly is created, the estimator attaches it to all four walls. It is the estimator's responsibility to ensure that the right assembly is attached to the appropriate design object. 
3) Create Estimate: After all the assemblies are attached, the estimator orders the estimating software to create the estimate and it automatically calculates the quantities for each item in the assembly. 
4) Manually Add Costs not Captured in Assembly : Finally, the estimator manually adds the costs for the additional production time for framing around the door opening.  This item could have been added to the cost assembly but then the estimator would have had to make two assemblies: one assembly for the wall without an opening and one assembly for the wall with the opening.  The estimator chose to make one assembly and manually add the costs for the opening.

Figure 2: Comparison of estimate generated using current tools and estimate 
generated by estimator to reflect the impact of design changes.

Designs change often and it is important to investigate how current tools help estimators to maintain estimates.  On the Sequus Project, the height of one wall changed from a height of 8' to 12.5' and a window was added.  Figure 2 shows that current tools reflect the change in design by calculating a new quantity.  Unfortunately, the changed quantity is not the only impact of this design change.  Changing the wall height and adding a window affected the metal stud and gypsum wallboard (GWB) in the following ways:
  • Resource composition: Rolling scaffolding was added to access the upper portion of the wall beyond a ten-foot height. 
  • Productivity rates: Productivity was reduced due to the hindered access to the upper portion of the wall from the rolling scaffolding. 
  • Quantity and cutting of GWB: The GWB are purchased as 4'x8' boards and must be cut to fit the size of the wall.  Since the wall height was increased to 12.5', additional boards had to be purchased and cut to fit the new height. 
  • Quantity and productivity of metal stud installation and cutting of GWB for Opening: The opening required additional metal stud framing with a different productivity rate.  Moreover, it required cutting of the GWB although the quantity purchased remained the same.

  •  

    Figure 3: Comparison of current process and Feature-based Estimating 
    process being developed as part of thsi research.

    This example illustrates the importance of identifying product features and properties and their effects on construction costs to support the maintenance of cost estimates as the design changes. However, current tools do not capture the estimator's rationale for relating product featuers and properties and their effects on cost information forcing estimators to manually create and maintain the cost information, as shown in the top portion of Figure 3. 

    To address this limitations, estimators need a framework to capture their rationale for relating product and cost information in a formalized and systematic way.  Moreover, estimators need a computer tool that utilizes the computer-interpretable representation of their rationale to assist them in identifying relevant product features and component properties in an IFC product model and their effects on construction costs. 

    My research addresses this problem by providing a computer-interpretable representation of estimator's rationale that enables the computer to assist estimators with the maintenance of cost information, as shown in the bottom portion of Figure 3. The Feature-based estimating process will allow estimators to enter their rationale easily in templates and the system will identify relevant cost information when creating estimates and notify estimators when cost information is no longer applicable in the case of design changes.

    Figure 4: IDEF view of Feature-based Estimating Process

    Figure 4 shows an IDEF view of the proposed research.  My objective is to provide a general framework for relating product and cost information that supports estimating for multiple domains and enables the reuse of estimating knowledge about component properties and product features and their effects. 
    Specifically, the objectives of this research are:

    1. To capture an estimator's rationale by formalizing product features and their relationships to construction cost information

    2  To formalize and test methods that analyze a product model to identify product features and relevant construction cost information

    * Laitinen, J. (1998). “Model Based Construction Process Management,” Ph. D. Thesis, Royal Institute of Technology. Stockholm, Sweden.