Finding technology that can eliminate workplace fatalities is the core mission of the National Safety Council's Work to Zero Initiative.

The organization's white paper, titled Using Computer Vision as a Risk Mitigation Tool, found that computer vision technology that uses a risk protection algorithm is capable of "accurate, consistent and automatic monitoring of worker personal protective equipment (PPE), risk identification, and the detection of violence and weapons in the workplace."

For industries such as manufacturing, warehousing and construction, software is available that is able to take current closed circuit television (CCTV) feeds that, when paired with artificial intelligence (AI), can rapidly review old footage to find and serve relevant post-incident images and clips. This information can be turned into dashboards for EHS professionals to use in determining root causes of workplace incidents.

The authors found that the most effective object detection computer vision systems are based on deep learning, which allows AI to identify key objects, including those that could be involved with incidents. This identification can help prevent future injuries.

Another key advantage of computer vision is the ability to monitor employees as they work, thus learning their habits and figuring out what specific activities could lead to incidents. The system is set up to continuously learn the association between behavior and incidents and turn them into tags that can be analyzed. 

For workplace injuries, computer vision can track and log data associated with these injuries and then use that to predict when injuries could occur.  The paper notes that this is especially useful for industries with material movement, such as logistics and warehouses, as well as those industries using heavy machinery such as manufacturing and construction. 

Other benefits of using computer vision in terms of safety include:

• Identifying when and where PPE is or isn't being used.
• Identifying safety hazards, in general
• Tracking the spread of infection
• Predicting the possibility of workplace violence.
• The system can be trained to recognize best practices during emergency situations.

As with any monitoring technology, there is a privacy concern. The paper points out that there are software packages, currently on the market, that are able to anonymize employee information and likeness to ensure privacy. 

The whitepaper goes into detail on these specific points, and concludes that, "computer vision technology, with the assistance of cloud computing, robust data synchronization systems and pre-trained A.I. and machine learning systems, is ready to deploy and deliver a real return on investment in terms of reducing the risk of workplace injury."