Everyone knows (or may have had experience being) the colleague or classmate who’s taken the easy route on an assignment knowing someone else will swoop in to save the day. New research suggests that the same phenomena, which psychologists call “social loafing,” may also apply when a teammate is made of software and hunks of metal.
A new study published in the journal Frontiers in Robotics and AI found that human workers checking manufacturing defects on circuit boards spotted fewer errors if they were told a competent robot had already reviewed the board once before. The findings suggest humans working alongside robots—an increasingly common site at major warehouses—may exhibit similar offloading of responsibilities as workers in all-human teams.
“Teamwork is a mixed blessing,” Technical University of Berlin of Berlin Research associate and paper author Dietlind Helene Cymek said. “Working together can motivate people to perform well but it can also lead to a loss of motivation because the individual contribution is not as visible.”
The Experiment
To conduct the experiment, the researchers asked participants to spot malfunctions or errors on circuit boards as part of a “simulated industrial defect-inspection task.” Half of the participants were told they were inspecting the board for the first time while another half were told the board had already been inspected by a robot, adorably named “Panda.”
The participants working with Panda were shown the robot and even heard it rumbling in another room while they inspected the circuit boards. Researchers provided all of the participants with 42 blurred images of the circuit boards which could only be unblurred by hovering a mouse over them. This nifty trick helped the researchers keep tabs on how long and thoroughly the participants parsed the circuit boards.
All of the workers involved in the study were then asked to rank their own effort levels following the task and comment on how responsible they felt for the work they just completed. It turns out that both the participants who thought they were working with Panda and those working independently spent about the same time inspecting the boards. They also both self-reported similar feelings of responsibility for the task.
But the proof was in the performance. The workers supposedly tasked with checking the robot’s first pass actually spotted fewer errors than those working independently. Despite the worse results, the participants working with Panda still believed they performed well. The researchers say that phenomenon resembles a “looking but not seeing” effect often observed in more traditional human social loafing cases. In other words, the humans appeared to trust the robot’s performance and overlook their own work.
Social loafing could lead to safety concerns
The study’s findings could serve as a cautionary tale for manufacturing and logistics companies rapidly moving to automate large warehouses and factories in the name of increased efficiency. Amazon, whose innovations in robotic logistics have helped it dominate e-commerce in the US, just revealed a suite of new artificial intelligence and robotics systems it plans to deploy at its fulfillment facilities to work alongside humans. Critics of Amazon’s push towards more automated warehouses, like the Strategic Organizing Center, have tried to link those changes to upticks in warehouse injury rates in recent years.
Those potential safety concerns were echoed by the recent study. Overconfidence in a robot’s performance could lead to a general lack of quality oversight or a potential overlooking of standard safety precautions. The researchers say those dangers potentially increase the longer work goes on.
“In longer shifts, when tasks are routine and the working environment offers little performance monitoring and feedback, the loss of motivation tends to be much greater,” Technische Universität of Berlin Professor Berlin Linda Onnasch, said. “In manufacturing in general, but especially in safety-related areas where double checking is common, this can have a negative impact on work outcomes.”