How AI could help make the workplace safer part two

(Note: This is the second in a three-part series on the use of artificial intelligence tools in the workplace. The first part is here)

Whether we call it machine learning or artificial intelligence, there are tools in the workplace now that can use video evidence and other tools to identify and flag a wide array of hazards in real-time, including wet floors, exposed wires, hairline cracks in machinery parts, or equipment left in dangerous positions. Its precision allows it to detect minute irregularities that might be imperceptible to the human eye. These kinds of tools can also verify the integrity and correct positioning of safety barriers, guardrails, or railings in the workplace, and ensure that workers maintain safe distances from operating machinery or hazardous equipment, and that all essential safety signs are correctly placed and clearly visible.

Some companies are using AI systems to do object detection and predictive analytics relating to workplace safety, including preventing collisions between pedestrians and heavy equipment such as forklifts. A properly trained AI system can recognize when a pedestrian enters a high-risk zone, automatically trigger warning signals for forklift operators, and even intelligently slow down forklifts in areas with high pedestrian traffic. Advanced AI systems can monitor workers’ movements to detect falls in real-time, sophisticated enough to differentiate between a worker rapidly descending a ladder and an actual fall. And some systems can identify when workers in isolated or high-risk zones are alone, enabling prompt assistance if an incident occurs, or monitor for unsafe work practices and deviations from established safety protocols.

AI’s proactive capabilities, driven by predictive analytics and real-time monitoring, can reduce the prevalence of workplace accidents and injuries, thereby enhancing overall safety. Research indicates that real-time safety indicator systems powered by AI can reduce workplace accidents by as much as 30%. The capacity to foresee and prevent accidents before they materialize directly translates into fewer injuries for employees and a corresponding reduction in workers’ compensation claims. AI integration also decreases operational downtime and mitigates the substantial costs historically associated with workplace injuries. Fewer workplace injuries directly translate into reduced absenteeism and lower workers’ compensation claims, yielding significant financial benefits. One estimate suggests that even a 10% reduction in workplace injuries could save US employers as much as $17 billion annually.  

Challenges, drawbacks, and potential ethical considerations

While the benefits of AI in workplace safety are compelling, its integration also presents significant challenges, drawbacks, and ethical considerations that demand careful attention and proactive management. The pervasive use of AI-powered surveillance tools, including cameras, sensors, and wearable devices, can lead to excessive monitoring of employees, raising significant concerns about potential infringements on individual privacy rights. AI systems inherently collect and store vast quantities of employee data, which poses substantial risks of data breaches and potential legal liabilities under various data privacy laws. This extensive data collection can foster unease among employees, leading to perceptions of the employer acting as “Big Brother” if the purpose and scope of monitoring are not transparently communicated and managed. 

While AI tools are designed to streamline health and safety operations, human oversight remains absolutely crucial for all decision-making processes, as AI is a decision-support tool, not a substitute for human judgment. Misinterpretations of AI outputs can lead to flawed safety actions if the AI-generated recommendations are not adequately contextualized or validated by human experts. In safety-critical environments, understanding why an AI system flagged a potential hazard or issued an alert is paramount. Without this transparency, human supervisors cannot effectively validate the AI’s output, make informed interventions, or conduct thorough incident investigations. Human decision-making must remain central, with AI serving as a decision-support tool rather than an autonomous decision-maker.

Organizations must obtain explicit and informed consent from individuals for data collection and provide clear, easily understandable privacy policies regarding data processing practices. Transparency about the AI technology’s purpose and limitations is crucial for building trust. A core principle is to collect and process only the personal data that is strictly necessary for the intended safety purpose, thereby reducing privacy risks. Privacy principles and safeguards should be incorporated as integral components from the earliest stages of AI system design and development. Organizations must adopt ethical AI principles and frameworks to ensure fairness, accountability, transparency, and respect for human rights. This includes conducting rigorous bias assessments before deploying AI features.

The third part of this series will wrap up some of the conclusions and lessons for employers considering using AI to monitor or improve workplace safety.

Interested in learning more about what a WCB employer representative can do for your group? You can connect with us on Facebook, Twitter, or Linkedin. Or you can reach us via email at [email protected], or you can contact us directly, during business hours, using our chat feature or by telephone at 1-844-377-9545.