Construct - Issue 49

In its desire to enhance safety and prevent accidents, FM Conway is using artificial intelligence (AI) to gain deeper insights into operational practices at its asphalt plants and depots. The technology is helping identify areas where safety systems can be improved to mitigate potentially risky situations and activities. The initiative stems from FM Conway’s Big Ten in 10 health and safety strategy, which leverages emerging technologies to enhance safety and eliminate risk from working practices. “When we plan work, we usually assume it is a linear sequence of activities, but the reality can be very different,” explains FM Conway SHEQ director Kevin Stevens. Safety professionals like Kevin use the term ‘practical drift’ to describe circumstances that result in actual performance varying from planned performance, which may impact on safety. “We wanted to find opportunities to address potential hazards before they result in an event,” he says. AI is an important tool because it can identify patterns of behaviour. An example is the reversing cameras installed on the business’s Liebherr loading shovels used to move materials around the asphalt plants. These machines spend around 50% of their time in reverse, so it is crucial that operators can manoeuvre just as safely in reverse as going forward. The cameras have been ‘trained’ to alert the operator and slow the machine down if they identify an object behind it. Crucially, they can tell the difference between a human and any other object, and if the camera detects a human within 8m, it issues a warning and automatically stops the machine. The system also records all these events, generating heat maps showing what is happening on the site and where the high risk areas are. “We can then look at why people are in that area, or why they are crossing at a hazardous location,” says Kevin. “Near misses usually require people to be there to spot something happening. With the cameras consistently picking up near misses, we get valuable information on where to focus our investment to reduce risks.”

with magnetic locking gates at all its asphalt plants. The gates are linked to traffic lights and only open when drivers have been signalled to stop. The system also projects red or green lines on the crossing surface to make it clear when and where it is safe to cross. AI CCTV cameras are also being installed at all FM Conway depots as ‘virtual supervisors’. They are trained to look for non-conformities and unauthorised access, so will pick up if someone is not wearing the correct PPE, if a vehicle is going in the wrong direction, or if someone falls over – which is vital for lone working. “All these incidents get reported back to a central system, helping us identify where the problem areas are,” explains Kevin. “We can use it to change processes and improve our safety systems in understanding why people make incorrect decisions. “Physical controls will always be required, but these technological tweaks help us identify practical drift.” At the Erith asphalt plant, for example, the cameras picked up staff checking stockpile levels while machinery was operating. “Through monitoring, we were able to identify a pattern of times when staff were checking stockpile levels during machinery operation. With this information, it was easy to change the process and eliminate human interventions in high-risk areas,” says Kevin. “For any risk professional, data is king, and this technology gives us the opportunity to get a detailed insight into what’s happening and improve our systems to provide better support to our people.”

As a result, the business has invested £1.2 million installing FHOSS walkways fitted

21