Stopping accidents before they happen key to data project

Big Data isn't just about helping retailers sell more products. Thanks to a project in development in Western Australia, it might also help reduce accidents on large mining and construction sites.

That's the aim of Justin Strharsky and his technology company Synaptor, which is taking incident logs and other reports generated on sites and analysing them to provide early warnings of risky behaviour or potential mishaps that may lead to injury.

As an indicator of the need for his company's technology, he cited Safe Work Australia statistics showing occupational illness and injuries cost the country $60.6 billion in 2008-09.

Synaptor's critical capability is its capacity to capture and manipulate live data, such as incoming reports of potential hazards like new potholes in roads or workers not wearing appropriate safety gear.

Synaptor analyses these reports and plots the data in real time on an interactive risk map, which is used to predict when and how an individual is most likely to be injured. It then generates an alert.

"One of the things we think is really key to preventing accidents is having the right information at the right time," Mr Strharsky said. "And part of the problem in industry now is that it is always at the wrong time. They are always responding to last month's data."

Mr Strharsky said Synaptor was in discussions with one site that already generated more than 40,000 observations about hazards and employee behaviour each month.

He said the accuracy of the system would improve as more sites provided data. There was also potential to bring in additional site data, such as shift length, swing length, and training records, as well as data from external sources that might have a bearing on safety, such as weather forecasts.

"Safety science has been doing the same thing for the past 60 years and has plateaued in its effectiveness," Mr Strharsky said.

Full story: smh.com.au/it-pro