Abstract
The construction industry is a core marker of social progress and a paramount driver
of developing national economies. The ongoing development of this sector has
become a foremost strategy of most countries. Over the past few years, notable
emphasis has been on adopting certified Sustainable Construction (SC) projects,
which have become crucial solutions to reducing CO2 emissions caused by traditional
construction and decreasing the consequences of climate change worldwide. Iraq is
recognised as highly active in developing and supporting this sector, especially SC,
by offering attractive investment opportunities and assigning a considerable public
budget. Nonetheless, the construction sector in developing nations, including Iraq, is
notably susceptible to challenges during COVID-19 that impede project undertaking.
Moreover, SC projects confront further challenges which hinder their execution in the
region. Effective risk assessment presents several advantages, such as determining,
evaluating and reducing risk factors, faster risk predictions with high accuracy,
makes better-informed decisions, improves chances of on-time project
completion within the allocated budget, and improves worker safety. Therefore,
companies must use analytical models that reduce time, address massive databases,
and make accurate predictions.
The aim of this research in the context of Iraq is to develop a risk assessment
conceptual framework that enables effective management, enhanced performance in
sustainable construction projects, and supports the industry throughout the life cycle
of projects. Additionally, five objectives were formulated in this research. To
accomplish the aim, a mixed-method approach was applied to collect and analyse,
including qualitative and quantitative approaches. Initially, an extensive literature
review was carried out to determine the knowledge gap and present a justification for
this investigation. This was followed by 1) three groups of semi-structured interviews,
2) focus group discussions, and 3) four groups of surveys. The interviews, focus
group and surveys were piloted, and all insightful comments were considered and
included to enhance their usability. Moreover, multiple approaches were applied to
analyse the collected qualitative data, including the probability and impact matrix,
content analysis and NVivo software (version 12) to determine the patterns, themes,
and classifications that arose. For the quantitative data analysis, multiple analytic
iii
approaches were employed, such as Failure Mode and Effect Analysis (FMEA),
Analytical Hierarchy Process (AHP), Artificial Neural Network (ANN), and Fuzzy
Inference System (FIS). Descriptive statistics, including sum, mean, and percentage,
were also used in this research. Finally, the Absolute Percentage Deviation (APD)
was employed to validate the generated ANN-Multi Layer Perceptron model.
The study findings revealed that, while traditional risk assessment techniques are
prevalent in the construction industry, AI-based risk assessment techniques can
improve risk assessment accuracy, reduce human errors, deal with massive
databases, present quicker risk predictions, provide real-time monitoring, and make
better-informed decisions. The results also indicated that the risk factors ranked by the
ANN approach and the mean method were strongly similar. Conversely, the risk
factors ranked by the AHP approach and mean method were significantly different.
This would prove the effectiveness of ANN-based risk assessment in predicting and
evaluating risk factors, analysing a massive database, and reducing human errors.
Additionally, the results revealed that COVID-19 has considerably impacted
construction management processes in Iraq, and the study determined the top five risk
factors that affected Iraq's construction projects, namely: (1) commitment to safety and
health recommendations; (2) risk management procedures; (3) equipment delivery
delays; (4) worker acceptance of the COVID-19 vaccination; and (5) increases to price
material. Furthermore, the results revealed the substantial risk factors that affected SC
projects in Iraq, including: (1) lack of specialists and professionals in SC management,
(2) inaccurate sustainable design information, (3) the need for a corresponding SC
contract, (4) poor cost estimation of SC, and (5) high initial SC cost.
This study contributes to knowledge by providing a robust ANN-based risk assessment
conceptual framework that analyses multiple variables and factors across the project
life cycle, providing a comprehensive and dynamic view of risks and their potential
impact. The framework also offers possibilities for effective, high-performance, and
quality outcomes, enhances predictive accuracy, provides real-time monitoring, and
enables data-driven decision-making. Finally, the study provides an insightful picture
of the substantial risk factors related to Iraq's construction sector, including SC
projects, along with practical policies and measures to reduce these risk factors'
impact.