Visual AI Saves Billions in HSE


Vol. 7 No. 3 // 2023

Business intelligence, financial, copyright Rikkyal via Creative Market

Examine the benefits of obtaining visual AI solutions that work 24/7 to keep workers safe

Artificial Intelligence (AI) has been making headlines for decades. It is a fascinating next-generation technology that has already created a massive paradigm shift worldwide. AI already has and will continue to alter the workplace, but its most significant impact, in my opinion, will be to reduce workplace injuries and fatalities.

Modified Image – Industry Workers, copyright biancoblue via Freepik


Workplace injuries and fatalities have long-lasting impacts on both individuals and organizations, emphasizing the critical importance of prioritizing safety in the workplace. In addition to physical injuries, mental health issues, such as stress, anxiety, and depression, can also arise from workplace incidents.

Heinrich’s Triangle Theory, developed by H.W. Heinrich, provides a framework for understanding the relationship between workplace incidents, near misses, and the presence of underlying factors contributing to accidents.

AI can play a pivotal role in breaking the chain of incidents outlined in Heinrich’s Triangle. Through advanced data analysis and machine learning, AI algorithms can identify precursors and early warning signs of potential accidents, enabling organizations to intervene and address these underlying factors before they escalate.


Many HSE challenges, such as poor ergonomics, a lack of personal protection equipment (PPE) compliance, and unsafe vehicle management, contribute to workplace accidents. Workplace accidents can range from multiple “near misses” to fatal injuries. However, human factors are almost always involved in workplace incidents.

Jodie Sasse, from SparkCognition recently gave a very insightful webinar on how their Visual AI Advisor helps to increase workplace health and safety. She pointed out that humans have a biased memory and are “notoriously forgetful.” She also noted that employees could arrive at work fatigued and follow old habits. In addition, out of all the risks that organizations face, failure to follow procedures is the most common, accounting for 62% of all threats. Another challenge for HSE management is that every employee digests employee training differently; some employees quickly understand their work, whereas others require practice.

  1. There were 5,190 fatal work injuries in the US in 2021. (BLS)
  2. Slips, trips, and falls account for 26% of nonfatal injuries in the workplace. (CDC)
  3. Over 2.4 million workers a year require ER visits after work-related injuries. (CDC)
  4. $171 billion is the cost of work-related injuries in 2019. (National Safety Council)
  5. An average of 14 workers per day lose their life on the job in the US. (OSHA)
  6. In 2020 deaths associated with transportation were the leading cause of workplace deaths. (US Bureau of Labor Statistics)


There is an enormous demand to have a reliable HSE solution. Investors are reluctant to invest in companies with a poor health and safety environment. Similarly, employees refrain from working with organizations that fail to provide a secure working environment. Workplace safety is one of the top concerns when considering a job offer.

Sasse shared that investing in HSE is the best way to put all these fears to rest. Companies save money by implementing competent HSE AI models, which diminish the chances of workplace accidents. In fact, for every dollar a company invests, their return on their HSE investment is twofold. Sometimes, it is even more.

Visual AI utilizes computer vision technology to identify and alert relevant personnel when certain health and safety parameters are not met. Using advanced machine learning models, it can capture rich media content, such as still images and videos, and classify objects such as people, machines, assets, etc. The software can then work in unison with existing camera equipment to identify any anomalies or compliance issues in the workplace.

A good visual AI solution can detect faulty equipment in the work environment and it can ensure flawless compliance with PPE. For example, Visual AI Advisor, an HSE solution designed by SparkCognition, can use cameras to identify workers who do not use the correct equipment for their work environment, like hard hats, gloves, etc. Likewise, it can switch off machines when it detects movement or personnel nearby.

Applying a robust machine learning solution, like Visual AI Advisor, will also ensure that as time passes and more data is collected the solution improves and learns to detect dangerous patterns to alert personnel with higher predictability.

  • Visual AI Advisor’s advanced HSE model is cost-effective and integrates with existing systems.
  • It ensures round-the-clock monitoring for visual protection while complying with privacy requirements.
  • It boasts company-wide coverage so that they monitor entire workplaces.
  • It dramatically reduces paperwork, bureaucracy, and work interruptions.


Visual AI Advisor’s HSE solution can be rapidly deployed and integrated with existing systems and can support over 125 use cases. Let’s examine some examples below.

The first use case is reduction of ergonomic stress. This problem covers about 58% of all HSE costs. Wrong postures and overexertion can cause this problem and Visual AI Advisor mitigates this and by identifying bad postures and notifying management to suggest changes. It also can implement corrective behavioral training.

The second use case is human and machine interaction. Here, the solution can track the vehicle’s speed and activity pattern through heat maps. It can then issue safety alerts in case of any safety risk detected. For instance, the breach of safety speed by a factory worker operating a vehicle.

Another use case worth examining is that of PPE guidance. It is imperative to implement PPE compliance and reduce workplace accidents. Visual AI Advisor monitors workers and their working equipment in real time. The AI can easily detect improper compliance with PPE guidelines. It can also track and notify policy adjustments if necessary.

One of the most interesting use cases is that of tailgating. Key-card entry alone is sufficient for regulating authorized access. However, Sasse raised the often-overlooked issue of tailgating. As the name suggests, tailgating in this scenario is when a person with authorized access opens the door with their key card, and someone not permitted into the restricted area follows them into that specific area. Tailgating is not only a glaring issue in a globally accepted security mechanism, but it can also tremendously compromise a company’s security and the confidentiality of its operations. Visual AI solves this problem using cameras and facial recognition software to verify who enters a specific area.

Modified Image – Industry Workers via Freepik


New technology is usually met with heavy criticism. This criticism circles mostly around privacy and the notion of constantly being watched. Sasse addressed this concern in the SparkCognition webinar and remarked that humanity is usually rigid towards change, whether new, beneficial, or detrimental. Concerns like privacy can be addressed by blurring the faces of the workers but implementing this solution could save lives. Therefore, in favor of workplace safety, the corporate work environment must embrace this change with open arms. It will not only preserve human life, which is priceless, and make work far more efficient, but it will cut down on costs that companies spend on worker protection with unsatisfactory results.

Related Articles

Action Plan