How might robots and artificial intelligence (AI) impact organizational safety protocols?

There’s been an increasing trend in industrial robot use over the past few decades. In 2004, companies installed 97,000 robotic units, while projections estimate 518,000 robots will be installed in 2024. That same year, forecasts by Gartner suggest that 75% of all businesses and organizations will also use artificial intelligence (AI) in an operational capacity.

The Occupational Safety and Health Administration (OSHA) points out that one of the main benefits of robotics in the workplace is that they handle dangerous tasks and take people out of harm’s way. Through automation, these machines can handle loading and unloading, welding while working in hazardous environments and other situations that commonly lead to workplace injuries.

According to OSHA, most injuries involving AI-powered machinery take place during maintenance. Lockout tagout devices can help avoid dangerous situations during maintenance. A ball valve lockout device or an electrical plug lockout, for instance, can prevent the machinery from operating until installation or maintenance is completed.

Take a look at how robotics and AI can impact safety and new considerations that these systems bring to the equation.

Enhancing workplace safety with robots and AI

As OSHA points out, robots can handle tasks with a higher risk of injury. Since they use the same process each time without fatiguing, automated machinery eliminates risks associated with human error. Injuries often occur due to changes in a work process brought on by fatigue or distraction — things that can’t happen to robots.

AI can further improve safety by enhancing monitoring and sensing capabilities, continuously analyzing safety procedures and automating emergency responses.

AI-powered safety monitoring and analysis

Artificial intelligence software and sensors can monitor processes and employee activities and look for potential anomalies in operations or the environment. Assessing this data in real time means the ability to alert management immediately or shut down operations until safety personnel investigate the issue.

AI systems can also crunch huge amounts of data, offering insights into the overall safety performance of employees. These insights can lead to improvements in procedures or changes to equipment or the work environment.

Predictive maintenance and accident prevention

While a funny term, the internet of things (IoT) relies seriously on connected sensors to retrieve and analyze data from equipment and work systems. By constantly monitoring the condition of machinery, AI can give early warning signs about maintenance requirements based on readings or changes to performance. These early fixes can help a business avoid accidents due to equipment malfunctions.

AI-powered monitoring can assess performance too, allowing companies to perform maintenance and ensure maximum productivity from all their equipment.

Improved emergency response

Robotics have long played a role in emergency response. Robots can perform cleanup tasks or assess the situation in an unstable environment while safety experts control them remotely — another way to keep employees safe.

Meanwhile, sensor-connected AI software can assist with post-accident procedures. Consider a warehouse fire. Sensors can detect where the blaze is and direct employees to safe exit routes based on real-time data.

Even if there is no disaster, AI’s data analysis can help companies predict events and timelines during a potential emergency, then put that framework into play when responding to an actual crisis.

The need for new safety protocols

Robotics and AI can make workplaces safer, but they require adjustments to existing worker protection protocols.

One of the most important aspects of safety in a workplace with robotic equipment is to notify employees of potentially dangerous equipment operations. This effort may include placing signage or other markings to notify employees of potential dangers or using locks and physical barriers to keep people out of operational areas.

As OSHA pointed out, robotic maintenance and work within equipment operation areas are common causes of accidents. A comprehensive lockout tagout program with lockout tags, for visual cues, and safety padlocks can protect against this danger.

Finally, employees will need to practice AI-enhanced evacuation and emergency response procedures so that they know how to interact and respond to these systems if an actual crisis occurs.

Collaborative robot safety

Many robots are not 100% autonomous, since they perform specific tasks alongside humans. These so-called “cobots” are designed to work on more complex tasks that require human management instead of programmed repetition. For instance, cobots can assist with repairs that require a human for assessment and problem-solving.

Because cobots perform different tasks, companies should perform a risk assessment to define dangers associated with the system. They can then take steps to limit these hazards. Safety options can include installing a monitored stop system or sensors that detect humans within an unsafe range and shutting down the robot immediately.

AI bias and ethical considerations

Bias is an ongoing issue in AI. Software responsible for safety and operational monitoring within a company runs based on data collection and pre-programmed algorithms. It acts based on inputs and isn’t able to consider the ethical implications or bias of its decisions.

Bias could be as simple as sensors triggering a shutdown because the system does not recognize someone with mobility issues or failing to register someone because of skin tone, height, weight or other factors.

Other workplace issues include collecting data on private interactions or from activities unrelated to safety or performance.

AI systems require human assessment and testing to ensure they function the same for all staff members.

Employee training

The safety of robotic equipment and AI-powered systems also depends on employee awareness. Workers need training so that they know how to react to these new workplace systems. These efforts could be as simple as educating employees about signage and lockout tagout systems and running drills to ensure proper response to AI-powered emergency systems.

Those who operate, repair and maintain the equipment need additional training on how to safely work on automated equipment and properly use safety systems. Ideally, training and assessment of safety procedures occur before the new systems are used in daily operations.

Legal and regulatory frameworks

Legal and regulatory frameworks for the use of robotics and AI are complex. Just because there’s a robot involved doesn’t mean there’s not an impact on people. Basic laws, such as those requiring safety signage apply to the use of robots if contact between machines and employees could prove dangerous.

At the same time, issues related to privacy and fairness are not well defined, and regulations depend on definitions and classification of systems and precedents set by court decisions. Companies can abide by existing workplace safety and privacy laws and be proactive in ensuring their robotics and AI systems are thoroughly tested for safety and bias.

Continuous evaluation and improvement

AI, autonomous equipment and worker-controlled robots are evolving quickly with innovations and breakthroughs happening frequently. Something as simple as a new software update can alter the safety protocols for an automated system. Companies should constantly assess new products, systems and updates to see if they can improve safety.

Getting started with AI and robotics in safety

The increased use of AI and robotics is impacting a variety of industries and careers – and occupational safety is no different. When evaluating the impact AI and robotics could have on your operations, consider which areas of your operations would benefit from automation. Are there simple or repetitive tasks that pose significant safety risks? Are there gaps in your safety data that AI monitoring could fill? These are great places to start in the early stages of implementing AI into your safety program. Once implemented, safety managers can use increased data capture to assess safety performance and continue to make impactful changes that increase overall workplace safety.