Adrian Latupeirissa 80% seminar: Adopting Film Sound Design to Foster Sonic Expression in Social Robots through Interactive Sonification

Abstract

This compilation thesis investigates the role of sound design in social robotics, drawing inspiration from robot depictions in science-fiction films. It addresses the limitations of robots' expressive behavior by integrating principles from film sound design, seeking to improve human-robot interaction through expressive gestures and non-verbal sounds.

The thesis is structured into two main parts. The first part focuses on perceptual studies, exploring how people perceive non-verbal sounds displayed by the Pepper robot related to its movement. These studies highlighted preferences for more refined sound models, subtle sounds that blend with ambient sounds, and sound characteristics matching the robot's visual attributes. This part also resulted in a programming interface connecting the Pepper robot with sound production tools.

The second part focuses on a structured analysis of robot sounds in films, revealing three narrative themes related to robot sounds in films with implications for social robotics. The first theme involves sounds associated with the physical attributes of robots, encompassing sub-themes of sound linked to robot size, exposed mechanisms, build quality, and anthropomorphic traits. The second theme delves into sounds accentuating robots' internal workings, with sub-themes related to learning and decision-making processes. Lastly, the third theme revolves around sounds utilized in robots' interactions with other characters within the scenes.

Discussant

Tony Belpaeme , Professor at Ghent University and Visiting Professor of Cognitive Systems and Robotics at Plymouth University.