Skip to main content

COMPUTER VISION

Using AI for Workplace Safety by Leveraging Real-Time Video

AI for workplace safety

Workplace safety is a real problem across industries. According to a recent report, a whopping 374 million people suffer from workplace accidents each year. It’s a scary number, but luckily businesses can act to reduce risks.

For example, many companies already have closed-circuit television camera (CCTV) systems in their infrastructure for routine security and asset management. These very same units can be used for additional purposes, such as identifying potential hazards to help decrease the number of workplace injuries.

Such implementation of video analytics technology runs on the premise that a chance occurrence or two will likely form an incident, but frequent repetition can indicate a wider, worrisome pattern. And when it comes to employee health and safety, these negative patterns can be costly but preventable by studying and acting on video camera footage.

Safety Analytics in Asset-Heavy Industries

When AllGoVision, a video analytics software solution provider, first started out in the CCTV field, it focused on security and asset management but quickly realized the technology’s potential beyond security. According to Aji Anirudhan, Chief Sales and Marketing Officer for AllGoVision, its use in protecting employee health in all types of industries became quite clear.

Asset-heavy industries such as manufacturing and energy and utilities can especially benefit from video analytics to improve employee working conditions. In such industries, worker safety can be compromised in two fundamental ways: first, when the situation on the ground changes rapidly. This is when accidents such as burns happen because of inappropriate handling of hot metals or exposure to toxic fumes from oil or gas leaks. In addition, workers can compromise their own health by not following safety protocols closely. Not wearing the right personal protective equipment (PPE) in high-risk working conditions, for example, can increase the risk of injury.

Asset-heavy industries such as #manufacturing and #energy and utilities can especially benefit from video #analytics to improve employee working conditions. @AllGoVision via @insightdottech

Traditional approaches to worker safety have been fairly passive, Anirudhan explains. Employee health safety teams conduct risk assessments of various plausible harmful scenarios and develop training programs accordingly. “And whenever an incident happens, they investigate it and try to update their policy so they will minimize these accidents,” he says. While company-wide policies are making a dent in the number of workplace injuries, the problem is still significant. Which is why “any solution which provides insights to understand patterns of unsafe behavior can help prevent accidents,” Anirudhan adds.

AllGoVision works hard to be that solution. Especially in energy and utilities companies, which work with oil and gas, safety problems abound through noncompliance. Whether it’s related to drilling equipment on oil fields or transport and storage of fuel, there’s a strong potential for mistakes at every step of the way, according to Anirudhan.

Mechanisms for AI Workplace Safety

To avoid compromising on worker safety, AllGoVision leverages AI to analyze the on-ground situation in real time. CCTV systems livestream video data, which AI safety analytics software can evaluate. It catches violations or problem areas in real time and can alert frontline managers who can proactively attend to these challenges. It provides dashboards that can help understand the current state of safety protocol adherence and monitor progress according to improvement plans. Employees also want to be part of the solution, and such analytics facilitate discussions of workplace data without a “Big Brother” approach, Anirudhan says.

AllGoVision pays special attention to data privacy protocols, preserving only metadata and not saving individual worker footage. The company also works to make sure its models are bias-free, and employs consultants to ensure compliance with international data privacy regulations.

Experienced Safety Analytics Vendor

AllGoVision particularly shines because the company harnesses its extensive experience in video analytics to evaluate specific client situations and recommend comprehensive solutions that will deliver desired outcomes. The software’s plug-and-play format means it can integrate with existing infrastructure, adding a data layer to derive information from. The company works with systems integrators to integrate solutions into a larger video management package that they might deliver to clients.

Experience executing workplace safety protocols in a number of industries matters. “Because AI is democratized, pretty much anybody can access algorithms available on open-source and create video analytics. But extrapolating that to a production environment where the cameras are different, the lighting is different, and the expectations are different, is the most challenging part,” Anirudhan says. “That’s where our strength in actually being able to address safety and employee productivity in different verticals especially oil and gas, utilities, and manufacturing, especially come into play.”

Over the years AllGoVision has evolved with different generations of Intel® platforms, Anirudhan says. The company was one of the early adopters of OpenVINO as well, and uses Intel® Xeon® processors. “The cost of running an algorithm has significantly come down and that is a saving for our customers. We see that as a huge advantage of working with Intel,” Anirudhan says.

Evolution of AI Workplace Safety

Anirudhan is excited about the many use cases for AllGoVision AI for workplace safety, including in buildings and infrastructure to detect fire and smoke, or in crowd control. The application of AI for workplace safety is still in the nascent stage, providing a huge opportunity for AI-driven solutions that make a significant impact.

Gone are the days of rinse-and-repeat policy-driven implementations. “Customers are seeing a clear value in moving to a real-time proactive approach,” Anirudhan says. Expect monitoring of more parameters, including worker fatigue, in the future. “There will be more wearables, sensors, IoT devices added to the workplace, which will all add to different use cases of machine and people management,” he adds.

In addition, Anirudhan sees enormous potential in generative AI to address more complex use cases, especially those that involve human-machine interaction.

AllGoVision is working hard to address worrisome workplace safety statistics. “We can make a big social impact if we could have AI actually address some of these challenges,” Anirudhan says.
 

This article was edited by Christina Cardoza, Editorial Director for insight.tech.

About the Author

Poornima Apte is a trained engineer turned technology writer. Her specialties run a gamut of technical topics from engineering, AI, IoT, to automation, robotics, 5G, and cybersecurity. Poornima's original reporting on Indian Americans moving to India in the wake of the country's economic boom won her an award from the South Asian Journalists’ Association. Follow her on LinkedIn.

Profile Photo of Poornima Apte