For decades, workplace safety has been about reactively auditing the work environment to pass a "tick box" exercise. This has not only led to high and sometimes fatal costs to businesses, but also higher expenses, more losses and a general inability to improve safety. But we are seeing changes – workplace safety puts loss prevention up front as a target, leading to lower loss ratios not just in regard to profits but more importantly for human life.
First things first: Let's acknowledge that the auditor model does not work
Time and time again, studies have shown that workplace safety improves when you let business owners manage their own safety. The more involved owners, managers and the workers themselves are in monitoring safety measures, the higher the chances of success. In fact, empirical evidence shows that safety incidents are one-seventh as likely to happen with engaged worker-centric approaches. Conversely when third party auditors are involved, more often than not companies just put up an appearance of compliance to get through the audit.
The results are lose-lose, disengaged workers, expensive auditors and no inherent increase in safety.
See also: Seriously? Artificial Intelligence?
The smarter the device or building, the safer the worker
Today we have the tools to genuinely anticipate and prevent accidents, and one of the best tools is that thing everyone carries around these days, a mobile phone. The list of wearables that can bolster workplace safety is also growing longer every day as we progress toward an Internet of Things.
With a smartphone, workers can take a picture of any hazard (for example, an electrical fault) with augmented reality, and the GPS on the phone turns the hazard into a dynamic alert as opposed to being some static and often hidden document. So, even if it is not removed immediately, as it is unlikely to be in most cases, workers can be alerted as they approach it. (Note: The experts seem to call this contextual awareness.)
The smartphone is just the start; the building itself is now smarter, with sensors for temperature, smoke, moisture, electric current, humidity, noise, light measurements, etc. In the more industrial workplaces, helmets, wristbands and even gloves are being embedded with sensors, so they can send alerts to employees and their managers in real time, allowing them to take preventive measures if workers’ well-being is compromised or safety procedures are not being followed.
As safety data pours in, machine learning steps in to make the most of it
While augmented reality is great for short-term risk management, machine learning makes sense of all the safety data collected and helps in long-term risk management. Placing this data among financial, environmental, occupational and social data can result in a system that updates in real time and any time (and not just via third party audits) and gives users GPS coordinates, pictures and notes. For insurance companies, this combination of IoT data feeding into machine learning capability will help deliver more sophisticated risk prediction models and underwriting risk assessment tools than the industry has ever seen before.
See also: Digital Playbooks for Insurers (Part 4)
There is no need to hide things from this auditor
Because it is self-audit!! But what does this mean to an insurance company? First, it means the focus now moves to loss prevention and subsequently, and carriers will have to lower premiums for worker-centric safety management. The lowering in top line premium is offset by lower expenses in using safety auditors and lower claims, leading to a better underwriting profit. This is not far-fetched; we have already seen this on the personal side and on the auto side with telematics. The trick this time around is combining with other data sources and machine learning for insights, which most humans could not comprehend in a traditional underwriting scenario.
I’ll leave you with a sobering fact – 4,836 fatal work injuries were recorded in the U.S. in 2015 itself
. That’s 4,836 too many. It is time for insurers to lead the charge on eliminating (the right kind of eliminating) with worker-centric processes powered by augmented reality and machine learning.