Buildability & Constructability
Buildability & Constructability
Buildability & Constructability
2.1
Where: A = Area of the beam; X = Productivity rate; Y = Total man-hour; n = Total
number of known productivity rate; j = Floor number.
35
However, the use of the above approach for constructability evaluation has serious
limitations. The major weakness is that it is difficult to imagine a situation in
construction where productivity rates remain absolutely constant for every floor
(Malek, 1996). Touran (1988) also stated that the regression analysis was not
successful on the first few floors because there were too many variations.
2.6.2. The expert system approach
An expert system is an information system that follows human lines of reasoning,
expressed in rules (e.g., IF/THEN rules), to arrive at a conclusion from known facts
(Mallach, 1994). The approach is extensively used to solve the problems in the
construction industry, as well as in buildability reviews and evaluation. For instance,
Alkass, Jergeas and Tyler (1992) developed a PC-based prototype expert system,
called Constructability Assessment for a Detailed Design system to evaluate the
constructability of a design detail. Fischer (1991) also developed a prototype expert
system to aid designers in achieving a more constructable reinforced concrete
structural system.
However, the classical expert system techniques have their inherent drawbacks, which
include: they cannot solve the problems that involve our subconscious use of common
sense that does not map well into production rules of expert system techniques; their
domain of expertise is usually narrow; they are brittle at their limits; and, they may be
costly to develop, in terms of the total time of the human experts and other people
involved in the process of development. In addition to the above-mentioned
drawbacks, production rules in conventional expert systems are deemed to be either
applicable or not applicable. However, most practical buildability problems depend
36
heavily on intuitive thinking and professional expertise that usually have a large
variation of shades of gray as opposed to black and white colors (Malek, 1996). Thus,
the classical expert system techniques have their limitations in solving problems in the
buildability domain.
To overcome the inherent drawbacks of the traditional expert system technique, the
decision support system was developed by combining three techniques, namely QFD,
fuzzy set theories and fuzzy systems. QFD provides a framework to broaden the
expertise by integrating the different disciplines into the decision-making processes of
buildable designs. Fuzzy set theories were used to capture the common sense
involved in buildable designs. Fuzzy systems provide a more flexible, economical and
reliable way to utilize the knowledge and experience in buildable designs.
2.6.3. The artificial neural network approach
Artificial neural networks (ANNs) were developed to mimic the biological neural
systems in learning, association, and generalization from training patterns or data (Yu,
1996). Based on the fuzzy ANN approach, Yu and Skibniewski (1999a, 1999b)
developed a prototype computer system, named Construction Technology Selector, for
constructability analysis and improvement of construction technologies. The system
has two basic functions. Firstly, the system can automatically acquire the buildability
knowledge through the combination of fuzzy logic with the learning abilities of neural
networks and genetic algorithms. Secondly, the system can evaluate construction
technology alternatives through a neuro-fuzzy knowledge-based multi-criterion
decision support system.
37
ANNs, especially the neuro-fuzzy approach, are promising methods to solve
knowledge acquisition problems in buildability reviews and evaluation (Yu, 1996; Yu
and Skibniewski, 1999a). However, the methods have several limitations and these
include: ANNs are black-box methods; ANNs are prone to having overfitting problems
due to their typical, large parameter set to be estimated; tedious experiments and trial-
and-error procedures are often used in ANNs; ANNs usually require more data and
computer time for training; and ANNs are data-driven and model-free and hence they
may be too dependent on a particular sample observed (Zhang, Patuwo and Hu, 1998).
For constructability evaluation and reviews, the decision makers want to know not
only how buildable a project is, but also why it is not builable. However, ANNs fall
short in answering the latter question due to its drawbacks (Yu, 1996). Moreover,
ANNs are also difficult in providing a structured mechanism for buildability reviews.
2.6.4. The fuzzy approach
Fuzzy set theories, introduced by Zadeh (1965), are widely used to solve the ill-
structured design and construction problems as well as buildability problems. For
instance, Malek (1996) developed a constructability assessment model for selecting the
optimum construction system for project execution. Chao and Skibniewski (1998)
developed a fuzzy-logic based risk-incorporating approach to evaluating new
construction technology. Generally, the fuzzy approach is a feasible and promising
approach for buildability reviews and evaluation (Malek , 1996; Yu, 1996; Chao and
Skibniewski, 1998). In particular, the fuzzy approach can be used to develop a
structured approach for formal buildability reviews and evaluation. A further
description of fuzzy set theories is given in section 4.2.
38
2.6.5. The Buildable Design Appraisal System (BDAS)
The Singapore Buildable Design Appraisal System (BDAS) was developed by the then
Singapores Construction Industry Development Board (CIDB) with the help of
various government agencies, leading local and foreign contractors, consultants and
product manufacturers, and employed the productivity data inputs supplied by them
(CIDB, 1993). The three principles, namely, standardization, simplicity and single
integrated elements were adopted by the system for buildability assessment (CIDB,
1993). Depending on the extent to which the three principles are adopted, the appraisal
system awards a set of parameter values or indices for each type of structural and
architectural system and other buildable design details (Poh and Chen, 1998).
The appraisal system employs equation 2.2 to compute a buildable score.
BS = 50[ (As Ss)] + 30[ (Aw Sw)] + N 2.2
Where: As =Asa / Ast, and Aw =Awa / Awt; As = Percentage of total floor area using
a particular structural design; Ast = Total floor area which includes roof (projected
area) and basement area; Asa = Floor area using the particular structural design; Aw =
Percentage of total external and internal wall areas using the particular wall design;
Awt = Total wall area, excluding perimeter wall of the basement; all internal walls in
the basement are to be considered; Awa = External and internal wall areas using the
particular wall design; Ss = Labor saving index for structural design, Sw = Labor
saving index for external and internal wall design; N = Buildability Score for other
buildable design features.
The Buildability Score of a project, which may consist of more than one building, is
computed by equation 2.3.
39
( )
( )
Ast building
BS project=Sum of BS building
Ast project
2.3
Generally, the appraisal system provides a feasible quantitative method to assess the
potential impact of design on labor usage during the construction process. However,
the appraisal system was criticized for its internal shortcomings. These shortcomings
mainly include: the appraisal system does not consider any project-related and site-
related factors that can affect site productivity; the appraisal system also does not
consider unit construction costs (Poh and Chen, 1998).
2.7. Summary
In this chapter, the state-of-the-art buildability and constructability was described and
discussed. It is noted that the theme that runs through the various studies completed
was to provide tools, methods and ways to integrate construction knowledge and
relevant information into all phases (planning, procurement, design, construction and
use) of the project lifecycle in order to achieve optimum project goals and building
performance.
The research aims at developing a decision support system for buildable designs. The
review of the principles and guidelines for buildability implementation provides the
theoretical foundation to develop the knowledge management model for buildable
designs (KM-BD, see Section 5.4). The review of knowledge management in
buildability provides the ways for organizing the buildability knowledge and
information into a manageable format. The review of knowledge management also
reveals the suitable functions of the proposed system. An examination of buildability
40
reviews and evaluation shows the advantages and disadvantages of the various possible
approaches to developing the proposed system.