How do people like to interact with robots when navigating crowded environments? And what algorithms should roboticists use to program robots to interact with humans?
These are the questions that a team of mechanical engineers and computer scientists from the University of California San Diego sought to answer in a study recently presented at the ICRA 2024 conference in Japan.
“To our knowledge, this is the first study investigating robots that assess human perception of risk for intelligent decision-making in everyday settings,” said Amodh Suresh, first author of the study, who contributed his Obtained Ph.D. In the research group of Professor Sonia Martinez Diaz in the UC San Diego Department of Mechanical and Aerospace Engineering. He is now a postdoctoral researcher at the US Army Research Lab.
“We wanted to create a framework that would help us understand how humans feel vulnerable when interacting with robots,” said Angelique Taylor, the study's second author, who Received his Ph.D. in Professor Laurel Rick's research group in the Department of Computer Science and Engineering at UC San Diego. Taylor is now on the faculty at Cornell Tech in New York.
The team turned to models from behavioral economics. But they wanted to know which one to use. The study took place during a pandemic, so the researchers had to design an online experiment to get their answer.
Subjects – Large STEM undergraduate and graduate students played a game, in which they acted as Instacart shoppers. They had a choice of three different routes to reach the milk aisle in the grocery store. Each route can take anywhere from five to 20 minutes. Some routes will bring them close to people with COVID, including one with a severe case. The routes also had different levels of risk of someone coughing up COVID. The shortest route puts subjects in contact with the sickest people. But shoppers were rewarded for reaching their goal.
The researchers were surprised to find that people consistently underestimated in their survey responses their willingness to take the risk of being near shoppers affected by COVID-19. “People don't mind taking a risk if there's a reward in it,” Suresh said.
As a result, to program the robots to interact with humans, the researchers decided to rely on probability theory, a behavioral economics model developed by Daniel Kahneman, who in 2002 published his Won the Nobel Prize in Economics for the work. Gain compared to a reference point. In this framework, people experience more harm than good. So for example, people will choose to win $450 rather than bet on something that has a 50% chance of them winning $1100. So the subjects in the study focused on getting the reward of completing the task faster rather than weighing the potential risk of contracting COVID, which was certain.
The researchers also asked people how they would like robots to communicate their intentions. Responses included speech, gestures and touch screens.
Next, the researchers hope to conduct in-person studies with a more diverse group of subjects.