A novel algorithm could significantly enhance robot safety by increasing their awareness of human distraction and inattention.
Computerized simulations of packaging processes reveal collaborative environments where humans and robots coexist, with an algorithm designed to mitigate human error significantly boosting security by up to 80% and efficiency by as much as 38%, surpassing existing approaches.
The work is reported in
According to Washington State College’s College of Mechanical and Supplies Engineering Assistant Professor Mehdi Hosseinzadeh, the majority of daily accidents can be attributed to carelessness, with most cases stemming from human error. While robots operate precisely according to established protocols, humans often fail to adhere to these same guidelines. The most daunting challenge lies here.
Robots increasingly collaborate with people across various industries, frequently working alongside them. While humans and robots often coexist in various industries, the monotony of repetitive tasks can lead to decreased concentration and increased mistakes among individuals? Robots’ laptop packages usually include mechanisms that help them respond appropriately in case of an error. Experts say that algorithms prioritizing efficiency and security often overlook the evolving behavior of users they’re intended to serve, according to Hosseinzadeh.
To inform their robot development plan, researchers initially endeavored to quantify human error rates, including metrics such as frequency of ignoring or missing security alerts.
As he spoke, his words betrayed a mixture of fascination and bewilderment as he described how we had meticulously outlined the careless actions, only for the robot to attempt to comprehend the complex behavior of humans. “The concept of a carelessness scale is an innovative approach.” If everyone agrees on which individual is consistently inattentive, we would likely take concrete steps to address the issue.
When the robot detects reckless behavior, its algorithms adjust its interaction style accordingly, seeking to minimize the likelihood of an employee mistake or self-harm in the workplace. As a result, the robot might adapt its workflow to avoid encroaching on human responsibilities. The robot continually updates the level of carelessness and records any observed modifications.
Researchers simulated a packaging line comprising four workers and a robot using a PC-based model to examine their plan. Additionally, they investigated a virtual team meeting scenario where two individuals worked alongside a robotic collaborator.
“Notably, the aim is to render the algorithm more resilient to potential misuse by individuals who might not fully grasp its intricacies,” he said. “Our research findings suggest that the proposed framework demonstrates significant enhancements in both efficiency and security.”
Following a computer-aided simulation, the scientists intend to verify their findings through experiments involving real robots and human participants, with the ultimate goal of applying these insights in an academic setting. Moreover, they must endeavour to quantify and account for various human attributes influencing office productivity, including individual rationality, risk perception and propensity to procrastination.
The research was supported by the National Science Foundation (NSF). Co-authors on this examination were Bruno Sinopoli and Dr. Aaron F. Bobick from Washington College, St. Louis.
