How to use AI to promote a proactive safety culture

Introduction

It is said that we learn most from our mistakes. Fortunately most workplaces do not have many fatalities or serious injuries to learn from. A lot of workplaces can operate for weeks without even a minor reportable accident. 

As a result, organisations wanting to improve safety don’t wait for accidents. They proactively seek out near misses and hazards, seeing these as a positive opportunity to involve the workforce in a learning process.

However, near miss reporting systems are fraught with difficulties. How do you get people to report? How representative are the reports you get? How do you get middle managers and supervisors to ‘buy in’ to the culture that when a worker reports their own mistake they are dealt with fairly, rather than punitively? Having received the reports, how does an EHS manager find the resources to make sense of them all?

AI can provide some of the answers.

Promoting reporting

Your workers are your eyes and ears on the ground for reporting near misses. Forward thinking organisations have made incident and hazard reporting easier using mobile reporting systems. These allow people to take a photograph on a mobile device, add a few details and send to a manager for review. AI can make such systems even easier to use, and more effective, by recognising machinery, vehicles or tools in a photograph, and looking for patterns of problems amongst large numbers of reports.

Now imagine that every CCTV camera was another set of eyes. A set of eyes that never got tired or had to take a meal break. A set of eyes that consistently reported every time someone walked in a vehicle zone, or an obstacle was left in a walkway. AI vision enables CCTV cameras to detect defined types of hazardous situations or behaviours, and report each and every one. The data collected will allow you to see where problems are occurring, and set about trying to fix them – before they lead to injury.

Demonstrating a just culture

Once you have the data to show how often hazardous behaviours or situations occur, you have an opportunity to show the workforce that the organisation has a fair and just culture. The aim is not to use the AI reporting to identify who is ‘cutting corners’ or ‘breaking the rules.’ Very few workers want to get hurt, or to cause harm to anyone else, so if behaviours are not as imagined in your written procedures, you need to find the underlying causes. Are workers taking shortcuts across vehicle routes to save time because an extra job has been added to the schedule this week to meet customer demands? Has the recent recruitment drive led to new employees being unsure when and where to wear PPE? Have new procedures reduced the time available for housekeeping, resulting in more trip hazards? Find out why the hazards are occurring, and involve the workforce in creating a workplace where people can naturally do the right thing. 

Proactive improvements

When the only data you track is accidents, it is difficult in the short-term to show any benefits from safety initiatives. Near miss reporting can be influenced by factors other than the underlying safety. When you track near misses using AI vision you have more reliable data to identify when things are starting to slip – and when investment in safety is paying off. You can see from the data that non-compliance with PPE wearing increased when new workers were recruited, and reduced again once a programme of training and toolbox talks was introduced. You can show the value of including information about PPE earlier in the process, perhaps with supervisors given time to show recruits what they need on their first day, rather than relying on colleagues. 

Conclusion

AI won’t replace the need to provide workers with a way to report issues in the workplace, but it will provide better data about the significance of a problem. Is the omission of a hard hat a one-off, or is it common? Did someone walk across a vehicle path once this week, or was it happening several times a day? Instead of spending hours wading through near miss reports, the AI does a lot of the hard work for you, identifying patterns and trends. AI can provide insights on questions we didn’t even realise we should ask. For the EHS manager, this could result in less time spent on reports, and more time to talk with people about safety and health. To learn more about how Protex AI is using their workplace safety software to detect risk before an accident can occur, encouraging businesses to embrace a proactive safety culture, chat to one of our product experts here 👈🏼

‍