Safety Culture in Practice

April Choy, Analyst at Verdantix on Serious Risk, AI and Speaking Up

During our webinar on safety culture, participants asked how to recognise serious risk, support workers on changing sites and assess whether technology improves safety. Verdantix EHS and Quality Industry Analyst April Choy answers six questions from the discussion below. 

• What does SIF mean?
  

  April: "SIF stands for serious injuries and fatalities. These are the most severe outcomes of workplace incidents, including injuries that may prevent someone from returning to work. A low overall injury rate does not necessarily tell you whether the controls for your most serious hazards are working.

• How can construction sites use AI when conditions keep changing and cameras are impractical?

  April: "I would start with tools people can use where the work happens. Mobile EHS software can help teams report incidents, carry out inspections, manage permits and check whether contractors are qualified for high-risk work. Wearable devices may help identify signs of fatigue or other physical concerns, depending on the task and device.
  
  The value comes from what happens next. Someone must assess the signal in its site context, decide on an action and check that the action has been completed. Cameras are one option, but they are not essential to that process."

• How can organisations encourage everyone to speak up and stop unsafe work?

  April: "I cannot share further details about the specific stop-work programme mentioned in the webinar. Its central principle is that everyone should know they have the authority to pause work when something seems unsafe, regardless of their role. 
  
  Leaders need to explain how that authority works and support people when they use it. A respectful response and visible follow-up make it easier for the next person to speak up."

• How do you keep wearables reliable over time?

  April: "Treat them as equipment that supports a safety process, with clear ownership and a maintenance plan. Build pre-use checks, scheduled calibration where required, maintenance and replacement into existing EHS asset routines. 
  

  It is also useful to monitor battery life, connectivity and device health. Otherwise, a gap in the data could be mistaken for an absence of risk. Device management works best when it forms part of the operational routine, rather than a separate technology process."

• What should organisations consider under GDPR when wearables or video analytics process personal data?

  April: "Start during vendor selection. Ask how the technology protects personal data, who can access it and how long it is retained. Some video analytics providers offer features such as facial blurring, anonymisation or processing footage on the device. Those features can reduce exposure, but they do not establish GDPR compliance on their own. 
  

  The organisation still needs an appropriate lawful basis, clear purposes and transparency with workers. It should assess whether the planned monitoring requires a data protection impact assessment and complete one where the processing is likely to create a high risk to individuals’ rights and freedoms. 
  

  Worker trust matters too. Explain how the information will help prevent hazards and protect people, and be clear about the limits on its use."

• Which metrics best show whether these technologies create value?

  April: "I would look beyond incident counts and total recordable injury rates. Track whether teams identify potential SIF exposures, whether critical controls work as intended, and how quickly they resolve corrective actions.
  

  Then examine participation. Are frontline workers raising concerns? Do they receive timely feedback? Are reported issues closed and lessons shared? The strongest evidence is a change in working practices or operational decisions that addresses a serious risk. A larger volume of data, on its own, is not a safety outcome."