FACE Investigations Make Recommendations to Improve the Safety of New Types of Robots

U.S. companies are installing robots in record numbers (1). These include traditional industrial robots separated from human workers by cages and cells, as well as emerging robotics technologies that include robots designed to work alongside and in the same space as human workers.

From a health and safety perspective, the proliferation of robotics technologies across various industries can be viewed as both positive and potentially concerning. While new technologies may be used to remove a worker from hazardous tasks thus reducing injuries, working closely with robots such as in a warehouse or on a construction site, may pose different hazards. These issues are discussed in the blogs: NIOSH Presents: An Occupational Safety and Health Perspective on Robotics Applications in the Workplace and A Robot May Not Injure a Worker: Working safely with robots.

Documenting Risk to Workers

Surveillance is the foundation of public health. However, information about both the number and specific details of robot-related injuries to workers involving newer robotics technologies are not well-documented. Currently, there is a lack of standard classification codes for robot-related injuries, which makes it hard to identify the frequency of incidents. Additionally, worker injury data systems do not include detailed information on how a robot-related fatality or injury incident occurred. This information is critical to inform policy changes and interventions to improve worker safety and health.

One of the ways we are learning more about injuries and worker deaths related to emerging robotics technologies is through case-based information from investigations. Investigations conducted by NIOSH’s Fatality Assessment and Control Evaluation (FACE) Program, or one of seven state health or labor departments that have cooperative agreements with NIOSH to conduct surveillance, targeted investigations, and prevention activities at the state level using the FACE model, help identify factors that contribute to fatal injuries. This information is then used to develop comprehensive recommendations for preventing similar deaths (2). The Washington State FACE Program recently made prevention recommendations for two new types of robots based on reviewing circumstances associated with a worker death and incidents resulting in severe injuries to two workers. Details are summarized below.

Warehouse Worker Killed by Driverless Forklift

The Washington State FACE Program (WA FACE) investigated a worker death involving a driverless forklift, known as a laser guided vehicle (LGV), at a water bottling company (3). LGVs navigate around this warehouse using a system of vehicle-mounted lasers and reflectors positioned throughout the warehouse. Each LGV has safety sensors designed to detect objects or workers in the vehicle’s path. In addition to an audible alarm, the LGVs are designed to stop moving until an obstruction is removed. However, per the manufacturer, an emergency stop also needs to be initiated for safety before removing the obstacle. Without the emergency stop, the LGV resumes moving automatically after the obstruction is removed.

In 2015, a 45-year old working as a warehouse dock coordinator at the water bottling company heard an alarm sound on one of the LGVs indicating that its sensors detected an object in its path. Investigators believe he attempted to remove a piece of plastic that likely tore off of a pallet and became stuck to the forks of the LGV without putting the vehicle in emergency stop mode. He was likely bending or kneeling under the forks, but outside the path of the sensor. When the object was gone, the LGV resumed operation. The elevated forks came down, crushing him against the vehicle. The worker suffered severe injuries. He was transported to a local hospital where he was pronounced dead. For more information see, “Fatality Narrative: Warehouse Worker Crushed by Forks of Laser Guided Vehicle.”

To prevent similar tragedies, WA FACE noted the following recommendations for employers:

• Incorporate manufacturer safety requirements into written company safety procedures for automated guided industrial vehicles.
• Train workers about the specific hazards and safety requirements associated with automated guided industrial vehicles, such as LGVs.
• Emphasize that workers are expected to follow required safety procedures every time, and ensure compliance through periodic refresher training and spot checks.

Workers Injured Using Demolition Robots

WA FACE also documented two cases of injuries to workers sustained while using remote controlled demolition machines, or demolition robots (4).

In the first case, an operator was crushed between the robot and a wall. The worker was using a machine that was connected by a wire to a remote controller he wore on his waist. After repositioning the machine, he attempted to move the machine’s power cable and bumped the remote control against the machine, which caused the machine to move and pin him between the outrigger and a wall. The machine was not in emergency stop mode.

In the second case, a worker suffered broken bones and nerve damage in his foot when a machine he was using crushed his foot. The worker stood in a tight spot between an excavation wall and the machine with a breaker attachment to chip concrete. As he tried to apply more pressure on the tip of the breaker, the front outrigger used to stabilize the machine raised off the ground. The machine shifted forward and the outrigger came down on his foot. For more information see, “Construction Hazard Alert: Workers Severely Injured Using Demolition Robots.”

In its Hazard Alert, WA FACE outlined the following recommendations for preventing injuries when working with demolition robots:

• Prepare a job hazard analysis with operators for each new job to identify and control hazards. Use the manufacturer’s safety instructions to establish the risk zone for the specific machine, attachment, and task.
• Always stay outside the risk zone when the machine is in operation, and do not enter until the machine is put into emergency stop mode or de-energized.
• Consider using a proximity warning system, such as those based on radio frequency identification (RFID), to maintain a safe worker-to-machine distance.
• Train operators to manage power cables and to continually monitor the process for hazards and redefine the risk zone.
• Ensure operators always read and follow manufacturer’s provided safety instructions.
• Consider using a spotter to assist the operator.

These cases illustrate some of the potential risks to those working with or around robotics technologies. As the excitement about these technologies continues to grow, we need to better understand how to keep human workers safe. WA FACE continues to look for cases of robot-related injured workers by searching for keywords in their state worker injury data system and reports from their state OSHA field inspectors. Having a standard robot-related injury or risk classification system would help identify and monitor these cases.

Occupational Safety and Health Science to Guide Safe Use of Robotics Technologies

NIOSH’s Center for Occupational Robotics Research (CORR) was established to provide scientific leadership to guide the development and use of occupational robots that enhance worker, safety, health, and well-being. The Center is building a research portfolio that includes efforts to document risk to workers through surveillance and fatality investigations.

Robot-related fatalities are expected to be rare, but should they occur, voluntary anonymous investigations, such as those discussed in this blog, will help us learn from these tragic events. Please contact the NIOSH FACE Program if you are aware of a worker fatality associated with emerging robotics technologies. Please contact CORR should you be in interested in partnering on other types of research or outreach.

Todd Schoonover, PhD, is the principal investigator for the WA FACE program.

Christina Rappin is a research investigator with the WA FACE program.

Randy Clark is a safety and health specialist with the WA FACE program.

Stephanie Stevens, MA, is a health communication specialist with NIOSH.

Dawn Castillo, MPH, is the manager of NIOSH’s Center for Occupational Robotics Research and director of the Division of Safety Research.