Why Machine Learning is the Next Big Thing for Construction Safety

Josh Kanner

More than 20% of workplace fatalities in the U.S. occur on construction sites. On average, that means a lack of on-site construction safety kills about 1,000 people a year, while one in ten construction workers will experience an injury in the same time frame.

Many construction companies give lip service to improving these numbers, but few do as much to improve worker safety as Skanska does. Since 2003, Skanska has made it their goal not just to improve construction safety, but to achieve a truly Injury-Free Environment. One of the largest construction and development companies in the US, Skanska makes safety and risk management core missions of their organization.

Toward that end, they use an integration of Smartvid.io, Autodesk’s BIM 360, and machine learning to continuously improve their performance.


Skanska construction safety helmet



Construction sites generate tons of photo and video data on a daily basis. From shooting photos of an installation, to drones delivering data on site conditions, this data has to be managed somehow. Often, it is either manually tagged and filed, or it is used once and then forgotten.

Smartvid.io is an industrial photo and video analytics platform that changes how photo and video files are managed on the job site. It analyzes all incoming visual data and accompanying audio data and automatically labels the data based on content, making it easy to search and sort.

Additionally, using a proprietary machine learning platform, Smartvid.io SmartTags can detect common safety risks like missing hard hats or gloves, giving teams an extra pair of eyes to help drive a positive safety culture on the project.

Since 2003, Skanska has made it their goal not just to improve construction safety, but to achieve a truly Injury-Free Environment


Integrating Autodesk BIM 360

Skanska uses Autodesk BIM 360 for construction document management, quality control, and other field activities. As a result, they manage a large quantity of photo, video, and audio data through BIM 360. Rather than loading data twice into two applications, Skanska uses Smartvid.io’s integration with BIM 360 to save time and improve results.

The integration automatically syncs all photo and video information between the two applications, giving Smartvid.io access to visual and audio information uploaded into BIM 360, and categorizes it for search. Then, Smartvid.io’s SmartTagging engine automatically tags potential safety hazards and notifies appropriate personnel when hazards are identified.


Watch: Talking Construction Technology and BIM 360 Forge Integrations with Smartvid.io


Machine Learning

Inside Smartvid.io is a proprietary machine learning technology called VINNIE. VINNIE uses a deep learning model to analyze vision and speech and develop a system tailored to the needs of construction. Because VINNIE is built into Smartvid.io, it learns from every new project. When personnel at Skanska interact with photos and videos, adding tags or categorizing material manually or interacting with it in other ways, VINNIE learns and gets better at its SmartTagging and at identifying safety risks.


Toward The Injury-Free Job Site




Skanska’s vision of an injury-free job site means they are constantly seeking new ways to move them closer to that goal. The combination of Smartvid.io with its machine learning and Autodesk BIM 360 for file management has led to major benefits for them, including:

  • Administrative time saved by automatic aggregation and indexing of visual data
  • Better job site risk analysis
  • Better collaboration thanks to ability to search and share dynamic links
  • Streamlined and improved safety processes

In the next phase, Skanska plans to sync more BIM 360 projects to Smartvid.io and increase the amount of video captured during the safety inspection process. This will help teams establish increasingly better best practices and processes for construction safety and risk management, and create a company-wide data repository across their thousands of projects worldwide.


Read: The complete Skanska case study for improving construction safety


Next Steps for Construction Safety

If you're ready to make radical improvements to safety on your construction site, watch the webcast "Using Machine Learning on Jobsites to Reduce Risk and Improve Safety."

In this webcast, thought leaders from Skanska, Smartvid.io, Autodesk, and NVIDIA will share their unique insights about the future of machine learning and construction safety.


Register to Watch: Using Machine Learning on Jobsites to Reduce Risk and Improve Safety 



Previous Article
Everything You Need to Know About Machine Learning for Construction Safety
Everything You Need to Know About Machine Learning for Construction Safety

Safety has always been a problem in the construction industry. More than 5,000 workers died on job sites in...

Next Flipbook
[eBook] Improving Construction Safety with Mobile Technology
[eBook] Improving Construction Safety with Mobile Technology

In this guide, we cover how you can take positive steps toward improving your safety management program wit...

UNLOCK the Power of Construction Safety Through Technology

Read Now