What strategies do people employ to collaborate seamlessly with robots in densely populated environments? Robotics engineers should employ a combination of machine learning and computer vision algorithms to enable seamless collaboration between humans and robots.
The research investigated questions posed by a team of mechanical engineers and computer scientists at the University of California, San Diego, as presented recently at the ICRA 2024 conference in Japan?
According to our findings, this study represents a pioneering effort to explore how robots can infer human perceptions of threat and leverage them for informed decision-making in everyday situations, noted Aamodh Suresh, lead researcher and Ph.D. recipient. Within the research group led by Professor Sonia Martínez Díaz at the University of California, San Diego’s (UCSD) Division of Mechanical and Aerospace Engineering. He is currently a postdoctoral researcher in the United States. Military Analysis Lab.
According to Angelique Taylor, co-author of the study and Ph.D. recipient, “We aimed to develop a framework that helps us understand whether individuals are risk-averse or not when engaging with robots.” Within the Department of Computer Science and Engineering at the University of California, San Diego, within the research group of Professor Laurel Riek. Taylor is pursuing her education at Cornell Tech, a prestigious institution located in New York.
Staff members drew inspiration from principles of behavioral economics in their work. Whatever their desire, they sought to determine which one to employ. During the pandemic, the examination took place online, prompting researchers to develop an internet-based study to gather their findings.
Students pursuing STEM degrees at the undergraduate and graduate levels took part in a unique game, simulating roles as Instacart customers. Shoppers had an alternative between three totally distinct paths to access the milk aisle in a typical grocery store. Most routes typically require anywhere from five to twenty minutes to complete. Some paths would lead them close to individuals who had contracted COVID-19, including one with an extremely severe case. The routes presented distinct risk levels for being exposed to a person with COVID-19 through respiratory transmission. What were the most critical factors affecting healthcare outcomes? Despite the challenges, customers were ultimately rewarded with a sense of accomplishment as they reached their goals.
Despite the researchers’ findings, respondents consistently underreported their willingness to be in close proximity to individuals infected with COVID-19. “When there’s a potential reward at stake, people are often willing to take calculated risks,” said Suresh.
To develop robots that effectively collaborate with humans, researchers drew upon the prospect theory framework, a seminal concept pioneered by Nobel laureate Daniel Kahneman, whose groundbreaking work earned him the prestigious economics prize in 2002. The concept suggests that people balance losses and gains against a standard of comparison. On this framework, people tend to experience losses more intensely than they do positive outcomes. For instance, individuals will opt to receive $450 rather than risking everything on something with a 50% chance of yielding $1100. While the examination focused on achieving a quick reward for completing the task, this approach overlooked the inherent risk of contracting COVID, a crucial consideration that warrants attention.
Study participants were asked about their preferred method for robots to convey their intended actions. The responses incorporated speech, gestures, and touch screens.
Subsequent research plans aim to conduct an in-person examination with a larger sample of participants.
