Layered animatronic robot head mimics singing/talking movements

Researchers from Sejong University and the Korean Institute of Science and Technology (KIST) have developed an animatronic robot head called Ray to help improve social interactions.

From Superinnovators 10/07/24

Ray, the animatronic robot head. Credit: Kyu Min Park et al, IEEE Transactions on Robotics (2024)

The robot head mimics human talking and singing movements through a combination of layered 3D printing and a tendon-based actuation system.

Ray’s structure includes contoured layers that represent human facial features, such as the eyes, nose, mouth, and chin, enhancing its realism.

The robot’s lightweight design and separate tendon-based actuation allow for smooth, fast movements of the head.

An audio-driven motion generation module enables Ray to synchronise its head and mouth movements with audio input automatically.

In video footage released this month, Ray can be seen moving to Charlie Chaplin’s final speech from the Great Dictator film, Sam Ryder’s Tiny Riot song and a South Korean song called Tree.

Demonstration of facial tracking. Credit: Kyu Min Park et al, IEEE Transactions on Robotics (2024)

The vocal parts of the songs were isolated from the rest of the instruments to allow the audio-motion module to trigger accurately.

Ray could be utilised in various applications, including integration with a humanoid to produce a talking robot, robot singer, or master of ceremonies or as a standalone head to represent AI chatbots.

The researchers believe this work opens up new possibilities for minimally designed audio-animatronic robots, potentially reducing production and maintenance costs.

The robot was entirely built using 3D printing technology, simplifying its production process.

This research was published in Volume 40 of the IEEE Transactions on Robotics journal.

The team at Sejong University and KIST say Ray could help revolutionise human-robot interaction in social settings.