Fuzzy Logic

Modeling Construction Labour Productivity Using
Fuzzy Logic and Exploring the Use of Fuzzy Hybrid
Techniques
Aminah Robinson Fayek, Abraham Assefa Tsehayae
Department of Civil and Environmental Engineering
University of Alberta
Edmonton, Canada
aminah.robinson@ualberta.ca
Abstract– Recent trends indicate that fuzzy techniques (fuzzy set
theory, fuzzy logic, and fuzzy hybrid models) havefound increased
application in the construction domain, even more so in the last half
decade. This paper presents the application of fuzzy expert models
and fuzzy hybrid concepts in modeling construction labour
productivity, which is critical information for scheduling and
estimating construction projects. The fuzzy expert model addresses
both subjective and objective factors affectinglabour productivity of
two common industrial construction processes: rigging and welding
pipe. The resulting model matched highly with respect to linguistic
terms; however, the numerical match was low, indicating the need to
have fuzzy hybrid models to improve the predictive ability of the
fuzzy expert model. Further research is underway to combine the
strengths of fuzzy logic in addressingsubjective and linguistic
evaluations of labourer performance with the strengths of other
artificial intelligence methods, such as neural networks, in training
and calibrating the fuzzy model to properly address the context
variables, as well as the principal variables.
Keywords-fuzzy set theory; fuzzy and hybrid fuzzy expert models;
construction industry; labour productivity

I.INTRODUCTION

The construction industry is a vital part of many national
economies. In Canada for the last five years, construction, on
average, contributed to 6% of the Gross Domestic product and
provided employment for about 7% of the workforce [1-2].
Despite its significance and long history, the industry has failed
in adopting new technologies that have brought improvements
to other industries,such as manufacturing [3]. It could be
argued that the construction industry’s complex and uncertain
nature might have prohibited the successful transfer of such
technologies. One technology, fuzzy set theory/fuzzy logic,
with its applications in many engineering disciplines, has the
potential to counter this tendency, in part because its strength is
dealing with complexity and uncertainty.The construction
industry is naturally fragmented and transient, with
unpredictable demands, long production cycles and, unlike
many industries, exposure to environmental factors [4]. As a
result, project management often involves use of approximate
linguistic terms and requires quick decisions based on complex
systems, imprecise or unstructured decision variables, and high
degrees ofuncertainty caused by effects of demand for

completed product, budget, weather, productivity, resource
allocation, etc. Fuzzy set theory provides useful solutions in
situations which involve highly complex systems whose
behavior is not well understood and in which approximate but
quick decisions or solutions are justified. Such situations are
common in today’s increasingly sophisticatedconstruction
industry [5].
As stated in [6], construction management studies “have
applied a wide range of statistical methods of analysis such as
regression, probability functions, mathematical learning curves,
and stochastic techniques for simulation and/or optimization”.
However “as complexity in construction management problems
has increased, the use of artificial intelligence techniques hasbeen explored and experimented with since the early 1990s”
[6]. A review of published research works related to fuzzy set
theory, fuzzy logic, and hybrid fuzzy techniques (combinations
of fuzzy logic with other artificial intelligence techniques, such
as artificial neural networks) from 1996-2005 indicated that
fuzzy techniques are being increasingly applied in construction
management...