A new algorithm could enhance robot safety by making robots more aware of human inattentiveness.
In computerized simulations of packaging and assembly lines where humans and robots collaborate, this algorithm, designed to account for human carelessness, improved safety by up to 80% and efficiency by up to 38% compared to existing methods.
The research is detailed in *IEEE Transactions on Systems, Man, and Cybernetics: Systems*.
“Accidents due to carelessness happen frequently—most, unfortunately, due to human errors,” said lead author Mehdi Hosseinzadeh, an assistant professor at Washington State University’s School of Mechanical and Materials Engineering. “Robots act as programmed and follow rules, but humans often don’t, which presents a significant challenge.”
Robots working alongside people are increasingly common across various industries, where shared workspaces are required. However, repetitive and tedious tasks can lead to human errors due to loss of focus. Most current algorithms help robots react to mistakes as they occur, focusing on either efficiency or safety, but they don’t account for the variable behavior of humans, Hosseinzadeh noted.
To develop a solution, the researchers first quantified human carelessness by analyzing factors like how often a person ignores or misses a safety alert.
“We defined carelessness, and the robot observed the human’s behavior to understand it,” Hosseinzadeh explained. “The concept of a carelessness level is new. If we can identify which humans are inattentive, we can take action.”
Once the robot detects careless behavior, it adjusts its interactions with the human to reduce the likelihood of workplace errors or injuries. For instance, the robot might modify its task management to avoid interfering with the human. The robot continuously updates its assessment of carelessness based on observed changes.
The researchers tested their algorithm using a computer simulation of a packaging line with four people and a robot, as well as a simulated collaborative assembly line where two humans worked alongside a robot.
Press release – Washington State University
