当COVID-19今年早些时候首次在美国遍布美国的大流行时，建筑工地看到了一系列新的安全程序，以确保在不随便传播病毒的情况下继续进行基本工作。建筑的安全文化适应了强制性的面具，社交距离和现场工人的手套。

Safety personnel try to enforce these rules, but measuring compliance can be tricky. That’s when AI-based image analysis firm Smartvid.io turned its system to more than just traditional analysis. Its AI tool can spot gloves and masks, and it defines social distancing as 6 ft between workers or groups of less than 10. Able to scan through thousands of videos and images of the site to categorize exactly what is going on, Smartvid’s AI soon was able to see trends among over 40,000 construction workers nationally.

“When we first started running analytics, there were marked differences in different regions, particularly for social distancing and mask-wearing,” says Josh Kanner, Smartvid.io founder and CEO. “Now six months later, all regions of the country are at a pretty high level of COVID compliance and behavior.”

While it is only a rough correlation, Smartvid.io’s regional trends for pandemic safety compliance on site do seem to follow outbreaks of the virus. The northeastern U.S. had seen fairly high compliance by early summer, while jobsites in southern states didn’t see a rise in social distancing until the late summer months.

坎纳说：“数据表明，所有地区都以积极的方式趋势，并以相当一致，高水平的社会距离和戴面具的融合。”“在这一点上，[美国]的所有地区似乎都在认真对待它。这与我们轶事所听到的一致。毕竟，如果您爆发了疫情，就会关闭。”

smartvid.ioinitially developed and deployed its COVID-related safety observations as a side project in the spring (ENR.com 4/3/20), but has since linked it into the general safety reports that the AI system generates, Kanner says.

The company also recently completed a funding round, raising $5 million. One of its latest investors is the Sony Innovation Fund, the investment arm of technology giant Sony. According to SIF, this is its first investment in an AEC-related company, although it has previously invested in industrial AI technologies.