Could a new tool save lives on construction sites?

A research team from the University of South Australia (UniSA) has developed a knowledge graph that can predict hazards on construction sites, replacing traditionally manual and time-intensive processes.

The job hazard analysis (JHA) tool aims to significantly reduce workplace accidents and improve safety in the high-risk construction industry – where 60,000 workers are killed each year worldwide, accounting for nearly 20 per cent of global occupational deaths.

UniSA Construction Management lecturer Dr Sonali Pandithawatta, who led the study, says potential job hazards rely heavily on safety personnel identifying risks and control measures, a process that is prone to inefficiencies and human error.

When designing the tool, the researchers used data from incident reports and experts, and integrated information such as weather, job steps, hazards, and preventive measures. Over 100 documents were analysed, with input sought from 18 industry experts to build the tool.

Co-author of the research, UniSA Construction and Project Management Professor Rameez Rameezdeen, says the model demonstrates more than 90 per cent accuracy. It is capable of analysing both primary and secondary hazards, the weather, workplace proximity and atmospheric hazards in real-time.

The team is now hoping to incorporate other risk factors into the research such as human and managerial influences, and to integrate advanced machine learning techniques for broader use across other high-risk industries.