Do more robots really lead to fewer workplace accidents? A rise in injuries at Amazon’s automated distribution centers may complicate a common assumption

October 22, 2020

There is the perception that AI-enabled robots will become a primary fixture in future workplaces. Companies like Amazon have already embraced robots as part of their daily warehouse operations. Today, however, robots are present at only 1.3% of firms. As such, there is still a great deal of uncertainty over what regulations and liability could apply when a business’s robot injures an employee.

The number of Amazon’s warehouse robots increased from 15,000 in 2014 to 200,000 in 2019. Though these robots have enhanced productivity, they, unfortunately, have come at a cost: “[i]n 2019, Amazon fulfillment centers recorded 14,000 serious injuries . . . . The overall rate of 7.7 serious injuries per 100 employees was 33% higher than in 2016 and nearly double the most recent industry standard.” Amazon has thus far not released specific information on the causes of such injuries. Still, an article by Reveal describes the company’s injury rates as “especially acute at robotic facilities[,]” despite Amazon’s claims that robots make work safer. The Reveal article points to increased worker isolation, “repetitive motions” in tasks, and “higher productivity expectations” that were “more than double” previous ones as consequences of the increased robotics use. In 2018 and 2019, Amazon was on the “Dirty Dozen” list for dangerous companies.

For all of their possible issues, robots have the capacity to not only make work more efficient, but safer too. For example, Robotics Business Review notes robots allow managers to “redeploy employees to more value-added positions[,] can take over “repetitive or dangerous jobs[,]” and provide increased strength to their human users. The same article also cites research that indicates human-robot collaboration results in 85 percent more productivity. An article in Forbes (funded by Mitsubishi Heavy Industries) argues that, especially in the time of COVID-19, robots can reduce safety issues from human activity while enhancing efficiency. Businesses can even use technology like robotic limbs and self-driving vehicles to provide work opportunities to individuals with disabilities who otherwise would not be able to perform certain tasks.

When an industrial robot does, however, cause harm, one question is what Occupational Safety and Health Administration standards could apply. Even today, OSHA has yet to adopt “standards for the robotics industry” beyond the Guidelines for Robotics Safety it released in 1987. OSHA can, however, use several existing rules to counter safety risks from robots in the workplace. For instance, OSHA standard 1910.212 relates to broader machine “operator and other employee[]” safety. OSHA also has the power to issue citations for when an employer has not performed a risk assessment on their robotics system. The Robotics Industries Association reports “OSHA’s view on robot safety is that if the employer is meeting the requirements of the consensus standards, specifically ANSI/RIA R15.06 – Safety Requirements for Robots and Robot Systems, then there will not be any issues.” Lastly, OSHA may also issue citations for the human side effects of increasing workplace automation. The Reveal article noted that “[a]s far back as December 2015 . . . . OSHA called out Amazon for conditions that resulted from the robotics innovation[.]” Specifically, that “‘[t]he company exposed employees to ergonomic risk factors including stress from repeated bending at the waist and repeated exertions, and standing during entire shifts up to 10 hours, four days a week and sometimes including mandatory overtime shifts.’”

“Where the assistance of lawyers, insurance companies, and updated guidance on robotics may be helpful is in the case of businesses that single-mindedly pursue productivity without thought for how robotics use could impact their employees.”

Another legal issue is who pays when a robot causes an injury. Specifically, whether liability should be confined to the employer/operator or extended to the robot’s manufacturer as well. As for employer/operators of robots, a National Law Review post stated employers are generally immune from civil suits because of workers’ compensation laws. In 2018, an Amazon warehouse robot punctured a can of bear repellant, sending 24 Amazon workers to the hospital. It is unclear what litigation, if any, followed the incident. Robot manufacturers, however, are third parties and thus may be liable for their products. Other sources, though, suggest robot manufacturers are not responsible for many safety issues, rather, the employer-operators should be the primary drivers of robot safety compliance

Today, robots comprise only a tiny fraction of workplace systems. Similarly, robot-caused injuries are presumably among the rarer workplace accidents. It is also important to remember that robots have immense potential to benefit the larger economy through greater productivity and can even make employee tasks substantially safer (robotic exoskeletons, for example, that reduce “wear and tear” from repetitive tasks). Where the assistance of lawyers, insurance companies, and updated guidance on robotics may be helpful is in the case of businesses that single-mindedly pursue productivity without thought for how robotics use could impact their employees. The Reveal article on Amazon suggests this way of business can exacerbate overexertion and repetitive motion injuries, both of which are among the top ten causes of workplace injuries. Thus, the issues Amazon is currently dealing with may be indicative of future problems as more and more companies embrace the benefits of AI-enhanced robotics.

Dakota Lipscombe