Improving group interaction for robots: a new data-driven technique

New approach to improving robot interaction in conversational groups
Robots must be able both to interact in groups and with individuals in social environments such as hospitals, malls and other public places. The majority of existing robots perform better when they communicate with individuals than groups of humans.

Hooman Szafir and Daniel Hedayati, two researchers from the University of North Carolina Chapel Hill, developed a data-driven method that could improve robot communication with groups of people. This method is presented in a paper at the 2022 ACM/IEEE International Conference on Human-Robot Interaction.

Hooman Hedayati told TechXplore that being in a group conversation is easy for people, but difficult for robots. Imagine that you’re talking to a group of people and, whenever someone blinks, the person stops talking and asks you if you’re still there. A robot in a conversational group can cause this potentially irritating scenario.