Framework Proposition for Identification, Impact Analysis and Intervention on Causes of Delay in Construction Projects in the Greater Bay Area Based on Descriptive Analytics and Artificial Intelligence
Date: 2023-11-14
Degree: Doctoral Thesis
Programme: Doctor of Business Administration
Authors: Fung Ka Chun
Supervisors: Prof. Alexandre Lobo, University of Saint Joseph
Abstract:
Construction projects are often complex endeavours, fraught with countless potential obstacles and challenges that can cause delays. In the case of the Greater Bay Area, in China, delays in constructions can have particularly profound implications potentially impacting on the region’s ongoing infrastructure development initiatives and affect its economic competitiveness and liveability. Although various studies have investigated construction delays, they have typically adopted generalised approaches rather than focusing on specific regions or contexts. However, it is becoming increasingly recognised that delays are context-specific and multifaceted, requiring more nuanced, localised understandings. This work proposes a framework to effectively manage construction delaying events, which has become a problem in the Greater Bay Area construction industry, including Hong Kong and Macau. Research findings are conducted in multiple phases. Firstly, by analysing scholarly articles, journals, books, and other interrelated online resources to identify the state of the art of the construction delay causes. From this phase, eight categories of construction delays in the Greater Bay Area were identified: Management, Design, Labour, Material, Equipment, Financial, Contract, and External/ Environmental. Furthermore, to understand the effect of the eight kinds of mentioned delay, it is, by all means, necessary to identify the nature of the delay and to classify each specific delaying event within the three categories: “Culpable Events”, “Excusable Events” or “Compensable Events”. Since the nature of delay will determine which party is defaulting, and which party will indeed recover the time and be responsible for eventual cost reimbursements. Most importantly, all events leading to “Extension of Time” must be included, along with fixing a new project completion date to identify the commencement of the delaying period. In other words, the original completion date as stipulated in the Contract will not be applicable if “Extension of Time” events are not accounted for; therefore, the actual delay will commence after all the issues mentioned above are settled. To compose a framework for identifying the kind of construction delay occurring the most in the Greater Bay Area, a structured survey with 100 participants with different backgrounds and experiences in the construction industry was conducted, and the result presented that Material, Labour and Management are the main categories of construction delays happening the most in the Greater Bay Area. The proposed framework in this work provides potential solutions for resolving the three significant delay categories and the five other types, considering the classification of delaying events as culpable, excusable and compensable. This research also provided a comprehensive quantitative data analysis, including an Exploratory Data Analysis (EDA) phase, followed by analysis of sub-groups using inferential statistics based on Kruskal-Wallis one-way Anova for data characterisation. Finally, Artificial Intelligence algorithms were considered, such as supervised Clustering Centroid technique and predictive analysis based on CART Decision Trees for each category of delay causes, with the aim to obtain realistic and practical solutions to prevent, mitigate and resolve different sorts of construction delays. The proposed framework, together with the advanced data analytics tools intends to be a relevant approach to support Project Managers and Project Stakeholders to predict and manage delays in both short and long term.
