In the last couple of months, we’ve heard a lot from our customers about their need to monitor COVID-19 safety compliance to help reduce the risk of jobsite outbreaks and shut downs. In response, Smartvid.io released new COVID-19 safety capabilities that enable our AI (nicknamed “Vinnie) to detect whether workers on a jobsite are social distancing, wearing masks and other important KPIs.
Now, as part of our work with a soon-to-be-announced technology working group of similarly minded companies, we have released our first set of cross-company analytics showing regional compliance with these COVID-19 safety protocols. We’ve drawn this data from our own customer base, and it spans 35,399 workers across the country for the first 3 weeks of June.
Our goal in releasing this information is to give construction organizations data they can use to benchmark their own compliance against others within their region. By reinforcing positive COVID-19 behaviors, we can all help reduce the risk of outbreaks and project shut-downs.
As you can see from the image, compliance with wearing gloves is high all across the country, probably because people were already accustomed to wearing them as a part of their day-to-day PPE on a jobsite. In all regions, we saw more than 90% of people on jobsites wearing gloves.
But mask wearing isn’t that far behind. Here, Vinnie counts everyone wearing some kind of face covering, from an N95 mask to a bandana. The West, Midwest and Northeast all had more than 80% of people complying, with the South lagging a bit behind at 72%.
We all have a ways to go, however, to achieve compliance with social distancing. First, here’s how we define social distancing: individuals are standing at least six feet apart and there are no groups of 10 people or more. The South, which came in last in mask compliance, also lagged behind the other regions for social distancing, with 43% of the images containing people not complying with social distancing. In every other region, the rate of social distancing issues observed was much lower in the low 30s and high 20s.
It’s important to note that we don’t enable our customers to collect this data so they can “nail” people who aren’t wearing a mask or are standing too close to their coworkers. In fact, we feel so strongly about this that we specifically do not enable facial recognition technologies. At Smartvid.io, we believe that collecting good data is the top priority, because without good data, we don’t have a factual basis on which to make decisions. If people are worried about getting themselves and others in trouble, they’ll be much less likely to make good observational reports, which produces bad data.
Instead, the goal is to identify trends so organizations can take action to reduce the threat of COVID-19. For example, people may be standing in groups not because they are careless about safety, but because there’s a bottleneck on the jobsite that’s forcing them to congregate together. People aren’t wearing masks? Perhaps others can’t hear them with their masks on due to ambient noise. If the goal is to identify individuals for punishment, safety personnel won’t likely look deeper to understand the environmental factors that shape safety behavior and take measures to correct them. And that would be a huge mistake for everyone’s safety.
In that spirit, we hope you’ll find this data useful as you benchmark your own organization's COVID-19 safety compliance against that of your peers. Subscribe to our blog (head to the top of the sidebar) to be notified when the next set of data are available.
Want to gain a better understanding of how your organization is complying with mask wearing and social distancing? Get in touch and find out the risk on your projects.