Skip to content Skip to navigation

Early-phase Rapid Evaluation of Prefabrication Alternatives for Building Structures

Project Team

Martin Fischer, Forest Flager, Yan-Ping Wang


The problem:  Quantitatively evaluating prefabrication alternatives for building structures in early design is currently not possible in the AEC industry, because practitioners are unable to evaluate the tradeoffs between expected value and uncertainty of construction process outcomes.

The proposed solution:  Develop a novel Rapid Construction Evaluation method (RCE) that automates the evaluation of construction schedule duration under uncertaintly in order to inform design decisions.

The proposed research approach:  Leverage the fundamental unit of construction – the construction part – to model fabrication and erection processes.  Develop a parts graph data model to automate the evaluation of the construction schedule.  Collaborate with researchers in advanced manufacturing industries to gain insights into methods and metrics for evaluating ease of assembly.  Validate the model through design charrettes and case study applications on industry projects.

Project Background

Research Motivation

Evaluating alternative construction methods in early design is not possible in current engineering and construction management practice.  In general, project teams are only able to evaluate construction alternatives after major design decisions have already been made.  By contrast, the ability to influence project outcome is the greatest precisely where there is the least quantitative, construction-based decision support.  Figure 1 illustrates the inverse relationship between the ability to influence the project outcome and the ability to evaluate alternative construction methods, in the context of the Paulson-Macleamy Curve:

Industry Example

A set of “construction-critical criterion” – including but not limited to cost, schedule, structural performance and constructability – are required for the evaluation of the tradeoffs between construction alternatives.  Currently there are no available methods for systematically evaluating building structures for these construction-critical criteria in early design, leading to significant missed opportunities in project performance improvements.  For example, our preliminary results indicate that a single building structural scheme can vary in cost by 40% and schedule by 80% based on the selection of construction method (figure 2).  

Research Objectives

Therefore, the goal of this research is to develop a novel Rapid Construction Evaluation method (RCE) that automates the evaluation of total construction cost, total construction duration, and total duration uncertainty.

Expected Results

Our preliminary results are able to illustrate tradeoffs between the expected value of a project and its uncertainty (Fig. 3).  While these results are exciting, the models require further calibration based on construction site data in order to generate more credible results.


We hope to gain access to high-quality footage of construction processes for both cast-in-place and precast concrete, in order to calibrate the RCE model.  We believe a correctly calibrated model can inform designers and engineers of the tradeoffs between expected value and uncertainty pf cost and schedule when choosing which components to prefabricate in a concrete structure.

PROJECT UPDATE - September 2018

To model the cast-in-place and precast concrete construction processes at a high level of detail, we have applied the Tri-Constraint Method (TCM) optimization framework via the ALICE Technologies software platform on a series of case study projects.  Each case study was conducted in the form of a 3-5 day workshop in the CIFE iRoom, with construction professionals from industry providing construction objectives, variables and constraints which were then used to automatically generate construction schedules for review.  The average time to generate a new construction schedule was 10 minutes.

Case Study: Cast-in-Place Concrete Structure (Industry Partner: SKANSKA)

The cast-in-place case study was conducted with SKANSKA Poland on a 32-storey, 72000m2 office tower project.   Four alternative construction strategies were analyzed by combining two sequencing alternatives (linear vs. parallel) with two formwork alternative (rail-climbing system (RCS) vs. self-climbing system (ACS)) (fig.4 ).  

During the workshop, these construction alternatives were generated and reviewed every 10 minutes on average, with over 300 schedules generated.  Furthermore, the results of the optimization yielded significant improvements for each constrategy versus the contractor's original estimate (fig. 5).

Case Study: Precast Concrete Structure (Industry Partner: Clark Pacific)

The precast case study was conducted with Clark Pacific on a 6-storey retrospective parking lot project at California State University at Sacramento.   Again, four alternative construction strategies were evaluated.  These were (1) baseline billboard strategy, (2) balanced crew size, (3) hybrid strategy, and (4) floor by floor strategy.  The initial strategy required 1 day to calibrate, while the remaining three required a half-day each to develop.  Overall, 56 construction schedules were evaluated out of over 500 construction schedules generated.  In this case, the optimization results for the baseline billboard strategy closely matched Clark Pacific's own precast erection schedule because the controlling factor was the availability of a single mobile crane performing the erection (fig. 5). 

The other three strategies did not perform as well as the baseline, but were expected to improve with further calibration.  This exercise demonstrated to the project team the impact of automated scheduling and how it can enable the rapid evaluating alternative construction strategies at a high level of detail before the beginning of construction.

Discussion and Next Steps:

By the end of each workshop, the participants believed this approach of rapidly evaluating construction alternatives could help to influence major design and construction decisions during the pre-construction stage.  However, upon closer examination there were still sequences in the construction schedules generated that are either sub-optimal, infeasible, or highly uncertain especially when activities are modeled at a high level of detail.  Consequently, the comparison of construction strategies is also relatively uncertain, and cannot inform decision-making unless the differences between them are sufficiently large.  Therefore the next steps of this research are to (1) determine the gap between current schedule automation methods and the ability to generate a "feasible and locally optimal" construction schedule, and (2) develop an uncertainty-based method for augmenting TCM to address this gap.  The expected contribution would be a construction optimization method that generates results with lower levels of uncertainty such that alternative construction strategies can be more accurately compared.

Original Research Proposal


Funding Year: 
Stakeholder Categories: