Development of a Life-Cycle Evaluation and Prediction Model for Construction Project Safety Level Using Organizational, Project, and Environmental Leading Indicators
This page is approximately a 5 minute read
This page was published on
Prof. Jaewook Jeong, Associate Professor, Department of Safety Engineering, Seoul National University of Science and Technology (SeoulTech)
Louis Kumi, Postdoctoral Researcher, Department of Safety Engineering, Seoul National University of Science and Technology (SeoulTech)
Jiwon Hwang , MSc. Researcher, Department of Safety Engineering, Seoul National University of Science and Technology (SeoulTech)
Construction remains one of the world's most hazardous industries; more than one in five construction workers experienced harm between 2020-2022. Workers are regularly exposed to risks despite improvements in safety regulations and management practices. Construction sites are also constantly changing, with vastly different risks encountered during construction phases from excavation to construction to maintenance.
The problem that our research is aiming to address is that construction safety performance is often assessed using records of past incidents, rather than assessments of what might happen, and also doesn’t take into account how factors can influence safety differently as projects progress.
Our research aims to address this challenge by identifying reliable leading indicators that can help predict safety risks before accidents occur. By understanding how these indicators interact and change throughout a project's life cycle, we hope to support a more proactive and preventive approach to construction safety management.
First, we will conduct a systematic review of existing evidence, analysis of construction project data, and develop a predictive safety model. This will help us identify which leading indicators have the strongest evidence of being associated with construction safety outcomes, and where important knowledge gaps remain.
Next, we will analyse construction project data to investigate how three groups of factors – organisational, project, and environmental – influence safety, and use statistical models and machine learning to identify patterns. We will then develop and test a model that predicts changes in safety risk across different stages of construction.
These findings will be published in an open-access online dashboard and toolkit that will help safety professionals identify emerging risks, prioritize preventive measures and make better-informed decisions.
The primary beneficiaries of this research will be construction workers, particularly those exposed to hazardous activities and changing conditions on construction sites.
Construction companies and site managers will be able to use this research to identify early warning signs of unsafe conditions during complex projects and take preventative action.
Site supervisors, safety professionals, contractors and project owners can use the evidence to gain a clearer understanding of which safety measures are most important at different stages of a project.
Regulators and policymakers may use the findings to support more targeted inspections, risk-based oversight and evidence-informed safety policies.
Although the predictive model will initially be developed using construction data from South Korea, the wider evidence review will draw on international research. This will help identify findings that may be relevant to construction safety management in other countries and contexts.
A key contribution of this project is its focus on understanding and predicting construction safety throughout the entire project life cycle, rather than treating safety performance as a single, unchanging measure.
Existing research has examined many factors associated with construction accidents, including safety training, management practices and environmental conditions. However, evidence about how these factors work together, which provide the most reliable leading indicators, and how their importance changes across construction stages remains fragmented.
Our project aims to bring this evidence together through a systematic review, assessing which indicators are associated with safety outcomes, and the strength and reliability of the supporting research.
We will then examine these relationships using construction project data and develop predictive models that account for changing project conditions.
By connecting evidence synthesis, real-world data analysis and practical digital tools, the project seeks to bridge the gap between safety research and everyday decision-making. The resulting evidence could help establish a stronger foundation for proactive, data-driven construction safety management.
Through collaboration and knowledge sharing, our goal is to ensure that the research produces not only scientific knowledge but also practical tools that can contribute to safer construction projects. We hope to reach the people who can put the findings into practice: construction workers, site managers, contractors, safety consultants, policymakers and regulators.
We are also keen to engage with researchers, construction companies, safety professionals, regulators, industry associations and organizations working to improve occupational safety, and invite them to contribute evidence to a systematic review examining how leading indicators can help predict and prevent construction accidents.
We are seeking research and reports examining relationships between factors that may indicate future safety risks and actual or measured safety outcomes in construction.
We are particularly interested in evidence relating to:
We welcome peer-reviewed research, government and regulatory reports, industry evaluations, technical reports, research institute publications and other relevant evidence, including unpublished studies and reports not readily available through academic databases.
Evidence from different countries, construction sectors and project stages are encouraged. The review will consider English-language evidence published from 1990 onwards.
Please send relevant publications, reports, working papers or links to accessible materials to Prof. Jaewook Jeong at jaewook.jeong@seoultech.ac.kr by 31 March 2027
In this blog, Giulia Maistrello, Senior Analyst at RAND Europe introduces the findings from two new reports. These reports underpin the Global Safety Evidence Centre’s new evidence programme on safe work, exploring both the evidence needs of safety practitioners and the evidence currently available to support them.
Professor Milena Nikolova from the University of Groningen is using World Risk Poll data to analyse how changing technologies affect safety in the workplace across different industries and countries, informing more targeted safety interventions.
Experts discuss the factors that make some workers more vulnerable than others to violence and harassment.
