As the construction sector in Australia strives to improve safety, the use of intelligence gained through sensors embedded in personal protective equipment (PPE) is emerging as a critical area of opportunity.

The sector has made significant gains when it comes to overall levels of safety. According to data from Safe Work Australia, the number of worker compensation claims for serious injuries fell from 27.5 per thousand workers employed throughout the industry in 2001/02 to 17 per thousand in 2012/13. Over the 11 years to 2013/14, the number of fatalities per thousand workers employed throughout the sector dropped by 31 per cent.

Nevertheless,  with the sector accounting for just over one in 10 serious injuries and having accounted for 18 out of 115 total workplace fatalities around the nation in the nine months to September last year, the need for ongoing improvement is obvious.

In this context, the insertion of sensors within personal protective equipment is one area in which action is happening. In road construction, for example, researchers at Virginia Tech in the United States have developed a prototype for a new type of vest which incorporates radio sensors and GPS tracking on the inside that couple with connected vehicle technology to warn both workers and motorists of likely impending collisions.

In Australia, RMIT has developed a system whereby sensors can be embedded onto safety glasses, helmets or boots and can monitor and inform workers of impending danger – such as if PPE devices are not being worn correctly.

Finally, Laing O’Rourke is trialing a type of networking arrangement involving an array of sensors mounted onto a sweatband with a data collection unit which can be retrofitted onto an existing hard hat. The sensors monitor the temperature and heart rate of the wearer as well as the external temperature and humidity conditions. The system alerts workers if and when they may be suffering from heatstroke.

All this prompts questions surrounding the potential of the technology along with any downside risks arising out of its adoption.

Speaking about his company’s hard hat, Laing O’Rourke device engineering lead Dr. Rod Shepherd says the anticipated benefits are expected to be twofold. Most immediately, as mentioned above, the sensors in the hat send sound alerts and vibration to the worker in cases where the system senses that he or she may be in danger of heatstroke.

In cases where the worker in question is online (which is not always the case as the presence of significant volumes of metal on some sites may mean that propagating data over radio link is not advisable), the information is sent back along a localised radio network to a gateway and is then pushed back out to site managers and on-site health and safety representatives. In this way, the device alerts workers themselves and potentially site management about the problem and thus allows them to consider whether or not the worker in question should have a short break.

Beyond that, there is potential to use aggregated data collected through the system to improve practices by looking at types of alerts received and analysing trends over time.

“Ultimately…we want to move into a preemptive or prediction condition in terms of our on-site safety management,” Shepherd said.

“That effectively moves away from that immediate response trigger and looks more at big data collecting information over a period of time and then trying to use that to improve your planning on site. It may be that we have a couple of alert scenarios over several months, but really the huge advantage can be gained long term in guiding how we do things day to day.”

With regard to further applications on site, Shepherd says an area of opportunity revolves around ‘geofencing.’ This could involve cordoning off sections of the site which, for example, might have a fixed mobile crane or be in a particular area of high traffic, with workers being warned through the hat upon approaching such an area. As shown in the Virginia Tech example, meanwhile, sensors involving hats or vests could also be used to warn workers and plant or vehicle operators of impending collisions.

In terms of drawbacks and challenges, Shepherd says these revolve around both technical issues and personal issues at the worker level.

In terms of technical matters, it is imperative to conduct appropriate research prior to implementation and to ensure tolerance zones are set at conservative levels, he said. In the case of a location base device for alerting workers of potential collisions which was proven to be accurate to within eight to 10 metres, for example, tolerance levels would need to be set at no less than around 15 metres. This is especially the case as those wearing the hat may have a tendency to become complacent in the belief that their hat will warn them of any impending danger.

Design is also important, Shepherd says, as devices will not be worn unless they involve minimal impact in terms of weight or ease of movement.

Beyond that, there are considerations relating to privacy and potential fears on the part of workers about information gleaned from monitors being used in ways which could impact their prospects for ongoing employment, for instance through the discovery or heart or other problems. In this regard, Shepherd says clear rules and agreements surrounding who can access data and how it will be used are essential, as is the de-identification of data used for broader analytical purposes.

Throughout the sector, the push to embed technology into hard hats and PPE devices is on.

If done well, this could be the next area of evolution in safety on site.