[Paper Review] In the Arms of a Robot: Designing Autonomous Hugging Robots with Intra-Hug Gestures

Alexis E. Block, Hasti Seifi+3 more|arXiv (Cornell University)|Feb 20, 2022

Music Technology and Sound Studies4 citations

TL;DR

This paper presents HuggieBot 3.0, an autonomous hugging robot that detects and responds to human intra-hug gestures—such as squeezing, rubbing, patting, and holding—using real-time haptic and audio sensing. With 88% classification accuracy, the robot uses a probabilistic behavior algorithm to deliver responsive, varied, and proactive hugs, significantly enhancing user enjoyment, perceived naturalness, and emotional connection in user studies.

ABSTRACT

Hugs are complex affective interactions that often include gestures like squeezes. We present six new guidelines for designing interactive hugging robots, which we validate through two studies with our custom robot. To achieve autonomy, we investigated robot responses to four human intra-hug gestures: holding, rubbing, patting, and squeezing. Thirty-two users each exchanged and rated sixteen hugs with an experimenter-controlled HuggieBot 2.0. The robot's inflated torso's microphone and pressure sensor collected data of the subjects' demonstrations that were used to develop a perceptual algorithm that classifies user actions with 88\% accuracy. Users enjoyed robot squeezes, regardless of their performed action, they valued variety in the robot response, and they appreciated robot-initiated intra-hug gestures. From average user ratings, we created a probabilistic behavior algorithm that chooses robot responses in real time. We implemented improvements to the robot platform to create HuggieBot 3.0 and then validated its gesture perception system and behavior algorithm with sixteen users. The robot's responses and proactive gestures were greatly enjoyed. Users found the robot more natural, enjoyable, and intelligent in the last phase of the experiment than in the first. After the study, they felt more understood by the robot and thought robots were nicer to hug.

Motivation & Objective

• To design autonomous hugging robots that can perceive and respond to intra-hug gestures in real time, enhancing emotional and physical intimacy.
• To address the lack of affective, responsive hugging robots in human-robot interaction, particularly for individuals isolated due to physical separation.
• To validate the impact of robot-initiated gestures and adaptive responses on user perception of naturalness, enjoyment, and emotional understanding.
• To develop a robust, real-time perception and behavior system that enables autonomous, safe, and socially intelligent hugs.
• To explore the potential of hugging robots in supporting mental health and strengthening distant relationships through embodied social touch.

Proposed method

• The robot uses a pressure sensor and built-in microphone on its inflatable torso to detect intra-hug gestures in real time.
• A perceptual algorithm classifies user actions (squeezing, rubbing, patting, holding) with 88% accuracy using audio and pressure data.
• A probabilistic behavior algorithm selects robot responses based on average user ratings from a prior user study, ensuring variety and user preference.
• The system includes a robust release detection mechanism that avoids accidental hug termination by ignoring brief pressure drops during grip adjustments.
• HuggieBot 3.0 was validated in a two-phase user study with 16 participants, comparing robot behavior before and after algorithmic improvements.
• A mobile app (HuggieApp) enables remote users to send customized hugs with gesture and video customization, supporting long-term emotional connection.

Experimental results

Research questions

• RQ1 How can a robot autonomously detect and respond to intra-hug gestures such as squeezing, rubbing, patting, and holding?
• RQ2 What impact do robot-initiated intra-hug gestures have on user perception of naturalness, enjoyment, and emotional connection?
• RQ3 How does real-time adaptation and response variety affect user experience in autonomous hugging robots?
• RQ4 Can a robot’s ability to perceive and respond to intra-hug gestures improve perceived emotional understanding and social presence?
• RQ5 To what extent can a hugging robot support emotional well-being and maintain relationships across physical distance?

Key findings

• The robot’s perceptual system classified intra-hug gestures with 88% accuracy using combined audio and pressure sensing.
• Users rated robot-initiated intra-hug gestures as highly enjoyable and appreciated the variety in responses.
• After algorithmic improvements, users perceived the robot as significantly more natural, enjoyable, and intelligent in the final phase of the study.
• Users reported feeling more understood by the robot and believed robots were nicer to hug after interacting with HuggieBot 3.0.
• The robot’s response system reduced accidental hug terminations by ignoring brief pressure drops associated with grip adjustments.
• The HuggieApp enabled remote users to send customized hugs, with potential for long-term emotional support, including legacy hugs after a sender’s passing.