How AI-powered real-time video analytics can enhance occupational safety

According to a study by the International Labour Organization (ILO), over 2.3 million workers around the world fall victim to work-related accidents or diseases every year! This accounts for over 340 million occupational accidents annually. Globally, the losses caused to the employer in terms of lost workdays, interrupted production, training, health-care expenditure sums up to around 3.94 percent of the world’s annual GDP.

Occupational accidents are very common, but to improve occupational safety, The ILO has defined more than 40 standards and codes of practice for workers in different industries. This includes standards for workers in sectors such as construction and manufacturing to wear proper safety props and safety precautions. But quite often, workers do not adhere to the safety standards defined by the employee, resulting in fatalities in case of accidents and causing monetary loss to the employer.

The occupational safety of workers constitutes their physical and mental well being. Having a proper occupational safety mechanism ensures that the health of the employees is not compromised and also helps in the prevention of hazards at the workplace.

During a survey that was conducted by the American Society of Safety Engineers (ASSE), it was found that 98% of respondents had observed that the workers were not wearing safety equipment when they were supposed to. And 30% had witnessed this on numerous occasions which were found to be a huge problem. Having an effective surveillance system to track and monitor the workers to ensure their safety can go a long way in preventing fatalities.

On a surface level, appointing a person to oversee workers seems logical, but the recurring events of workplace accidents have proven that this method is not very effective. To ensure the safety of the workers, it is essential to monitor all workers continuously to assure they are wearing all safety props and following the safety guidelines. This is not something a human can possibly do. And that’s where AI-powered real-time video analytics can be helpful.

AI-powered real-time video analytics

Today, technology has advanced a lot. Computers can analyze a video more efficiently than humans (who are naturally limited by their attention span of merely 8 seconds). Real-time anomaly detection systems such as Emotyx can analyze footage from CCTV cameras and identify the occurrence of unusual activities, such as accidents, fires, leakages, etc. Wouldn’t it be better if there is a digital eye supervising the factories and production facilities and timely alert the security officials if it identifies workers who are not following safety measures? It is beyond human capability to observe and record every minute abnormality. Certain events specific to your workspace can go unnoticed. However, it is possible to create alerts with AI-powered real-time video analytics that help detect unusual behavior and help your organization rectify it to avoid further damage. 

AI-powered video analytics can analyze the video stream from CCTV cameras to identify objects, patterns, anomalies, etc. In this case, it can be used for;

• Safety Props Detection: To ensure the workers are wearing all the required safety props such as helmets, jackets, boots, etc
• Process monitoring: To ensure that the workers are following the required processes. Any deviation from the defined process can raise alerts and notify the supervisor.
• Machinery monitoring: Monitor and ensure that machinery is working and operated under safety conditions.
• Anomaly detection: Monitors the workplace for anomalous events like the occurrence of fire, leakages, spills, etc and notify the supervisor.
• Detect restricted area access of people standing or loitering, vehicles parking in handicapped zones or fire lanes exceeding a specified period.
• Analyze the operations of heave machinery and collaborate with forklift drivers and warehouse staff to give instructions effectively.
• Restricted zone detection of suspicious people and vehicles at production facilities and warehouses.
• Detection of irregularities in warehouses based on shipment type, positioning, and other factors.

Not only is video analytics an outstanding advancement in perimeter protection and loss prevention for warehouses and distribution centers, but it is also increasing in its importance and demand for use in other markets. Instead of having operators monitor multiple cameras, video analytics identifies behaviors of interest and immediately alerts the security operator to suspicious activity. Ultimately, this allows security operators to focus their attention on the most important activities and monitor the highest risk potential and enables them to make better decisions and react faster to events.