Login | Register

A Predictive Model for Scaffolding Manhours in Heavy Industrial Construction Projects: An application of machine learning

Title:

A Predictive Model for Scaffolding Manhours in Heavy Industrial Construction Projects: An application of machine learning

Siddappa, Kavana (2019) A Predictive Model for Scaffolding Manhours in Heavy Industrial Construction Projects: An application of machine learning. Masters thesis, Concordia University.

[thumbnail of Siddappa_MASc_S2019.pdf]
Preview
Text (application/pdf)
Siddappa_MASc_S2019.pdf - Accepted Version
2MB

Abstract

In cold countries like Canada, modular construction is widely adopted in heavy industrial construction projects due to weather uncertainties. To facilitate the construction processes, the temporary structures, especially scaffolding, are essential since it provides easy access for workers to carry out construction activities at different levels of the height and also ensures the safety of laborers. As indirect costs of projects, the scaffolding is estimated by 15-40% of project costs. Furthermore, according to the increase in the size of the projects, the scaffolding uses a larger amount of resources than estimated ones, which may cause budget overrun and schedule delay. However, due to the lack of systematic and scientific models to estimate scaffolding productivity, the heavy industrial company has difficulty to plan and allocate the resources for scaffold activities before construction. To overcome these challenges, this paper proposes a predictive model to estimate scaffolding productivity based on the historical scaffolding data of a heavy industrial project. The proposed model is developed based on the following steps: (i) identifying the key parameters (e.g. specific trades, work type, different scaffold methods, task times spent using scaffolds, and weights of the scaffolds) that influence the scaffolding manhours and project productivity; and (ii) developing the predictive models for scaffold manhours using machine learning algorithms including multiple linear regression, decision tree regression, random forest regression and artificial neural networks(ANN) . The accuracy of models have been measured with evaluation metrics which are mean absolute error (MAE) and root mean squared error (RMSE) and the R squared value. The findings reveal up to 90% accuracy for ANN models.

Divisions:Concordia University > Gina Cody School of Engineering and Computer Science > Building, Civil and Environmental Engineering
Item Type:Thesis (Masters)
Authors:Siddappa, Kavana
Institution:Concordia University
Degree Name:M.A. Sc.
Program:Building Engineering
Date:16 July 2019
Thesis Supervisor(s):Han, Sang Hyeok
ID Code:985585
Deposited By: Kavana Siddappa
Deposited On:05 Feb 2020 14:04
Last Modified:05 Feb 2020 14:04
All items in Spectrum are protected by copyright, with all rights reserved. The use of items is governed by Spectrum's terms of access.

Repository Staff Only: item control page

Downloads per month over past year

Research related to the current document (at the CORE website)
- Research related to the current document (at the CORE website)
Back to top Back to top