Until recently, AI insights were straight out of sci-fi. Today, in the medical field, robots alert doctors to risks during surgery; bankers use predictive data to warn clients before their spending habits lead to overdrawn accounts. We’ve seen firsthand how AI can analyze construction job site safety. Lately, however, we’ve been looking toward the future: one in which AI can identify and prevent construction risks long before they become a threat.
Looking backwards, moving forward
Every job site is different, from the individuals present to day-to-day weather conditions. As a result, there’s never been a straightforward way to quantify the benefits of recognizing and reducing risks. That’s why we analyzed a decade of data to learn how to apply predictive analytics to construction projects.
In collaboration with Suffolk, our team observed and reported safety risks as part of the “Safety Together” program, which fosters collaboration between Suffolk and subcontractors and encourages positive, informed conversations about safety. Then, we applied our AI engine to check for PPE compliance and observation-based risks across 10 years of Suffolk job sites. The result? Our AI engine, Vinnie, predicted one in five incidents with 80 percent accuracy.
Safety in numbers
Depending on the number of alerts the site manager opts into seeing, alerts can fire for an even greater percentage of the total incidents. Even if only 25 percent of these predicted incidents are ultimately avoided, a company with 50 annual projects will see 40 to 100 fewer incidents per year!
Every avoided incident represents the health and well being of an individual worker. From a monetary perspective, each incident costs approximately $36,000--and avoiding 40 to 100 per year results in saving $1.4 to $3.6 million. Without accounting for insurance premiums and bills, it’s clear that investing in a safer, more predictable work environment more than pays for itself.
For those in the construction industry, AI offers a unique perspective. Safety, quality, costs and schedules are the factors driving any project’s success. When construction managers take a closer, automated look at each, and make predictions based on real data, they go beyond the ability to reduce workers’ comp claims. They can save millions, control hazards and improve safety performance, all while building something great.