Talking Robots Learn to Manage Human Interruptions
Johns Hopkins computer scientists designed an interruption-handling system to facilitate more natural conversations with social robots
By Jaimie Patterson
Johns Hopkins University researchers have created a system that could make social robots more effective at detecting and managing user interruptions in real time based on a human speaker’s intent—a breakthrough for areas like health care and education where natural conversation is crucial. The team presented its work at this year’s Robotics: Science and Systems conference, held in Los Angeles June 21 to 25.
Despite all their advancements, state-of-the-art robotic systems still have difficulties handling user interruptions in real time—and often don’t understand why humans interrupt in the first place.
So Jiwon Moon, Engr ’25; computer science PhD students Shiye “Sally” Cao, Amama Mahmood, and Victor Nikhil Antony; and assistant professors Ziang Xiao, Anqi “Angie” Liu, and Chien-Ming Huang began by analyzing different types of human conversations like discussions, talk show interviews, and press briefings to identify how humans handle someone talking over them.
The researchers observed that interruptions can have different purposes, such as signaling understanding, assisting the speaker, seeking clarification, expressing disagreement, further developing the topic, or changing the subject. Likewise, those who were interrupted reacted in various ways: ignoring the interruption, acknowledging it but continuing, or yielding to the interrupter.
The team also recommends that future work explore non-verbal interruptions—like a user opening their mouth to speak without saying anything—and investigate interruption handling in longer or multi-session interactions with multiple users.