Development and comparative analysis of the fuzzy inference system-based construction labor productivity models

Abstract

The current practice to estimate construction labor productivity (CLP) lacks a systematic approach to measure and estimate it. Moreover, modeling CLP is challenging because, in addition to the realistic constraints of multiple factors, subjective assessments, low-quality data, and limited datasets, the complex relationship between multiple factors should also be considered simultaneously. Such challenges were addressed in this paper through developing, optimizing, and validating a series of CLP models. Appropriation of using artificial neural networks (ANNs) to model CLP has been previously tested. However, ANNs cannot incorporate linguistic assessment of qualitative influential factors, which is indispensable for CLP modeling. Therefore, the current paper also proposed two CLP modeling methods based on a fuzzy inference system (FIS). This study contributes to the construction engineering and management body of knowledge by proposing a modeling technique, capable of dealing with a combination of crisp and fuzzy input variables. Moreover, a comparative analysis helps researchers and practitioners choose the appropriate method according to the nature of their projects, and through its implementation, estimate the productivity with a level of accuracy and interpretability greater than what could be offered by previous techniques.

Keywords:

• Construction labor productivity
• artificial neural network (ANN)
• fuzzy inference system (FIS)
• adaptive neuro-fuzzy interface system (ANFIS)
• artificial intelligence
• qualitative variables

Disclosure statement

No potential conflict of interest was reported by the authors.

Date availability statements

Data generated or analyzed during the study are available from the corresponding author by request.