Building Social Companions

Overview

This line of work contributes to our understanding of how robotic products and services might function in real-world environments is my group’s work on long-term interaction with robotic social companions, focusing on family-robot interaction, conversational privacy, and ways of maintaining interaction over long periods (months).

Over the course of the last four years, we have iteratively designed a learning companion robot for middle schoolers, called Minnie, that aimed to socially augment the otherwise solitary activity of reading. Children read to Minnie, and the robot occasionally comments on the story, reflecting on characters and events and connecting the story to other stories they have read. The robot also helps the child choose books to read based on the child’s interests and meet reading goals. We have fabricated multiple copies of Minnie and carried out in-home evaluation studies, including a 14-day deployment study, which revealed that children established an emotional bond with the robot over the course of the 14 days and reported being more motivated to read and better understanding the content with the robot than with a paper-based reading activity. We also have conducted laboratory evaluations of how Minnie can help build interested in science reading, and we are now planning a summer-long deployment of Minnie with low-interest readers to better understand how interactions with and perceptions of the robot changes over a longer period.

Here is a short talk I gave at the 2022 Furhat Conference on this topic:

Research Team

In this work, I collaborate with Joseph Michaelis, David Shaffer, and Sarah Sebo.

The students who participate in this work include Bengisu Cagiltay and Nathan White.

Publications & Videos

The key publications that has resulted from this work include:

This work has also appeared in popular scientific news outlets, including:

Funding

This work is supported by NSF awards 1906854, 1822872, and 2247381.

NSF Logo