DE
Menü öffnen

Exploring the multimodality of Arts with the help of Robotics

What happens in the human mind when we listen to music and express our feelings through painting and movement? Can a robot interpret music to paintings in a similar way to humans? These are some of the questions we are exploring in this exciting project using mechatronics, art, AI, and workshops with children. 

Description of the project

Display of the painting robot
Credit: Franziska Fiolka

Within the framework of the Ligeti Center, this project aims at combining art and science. It consists in the development of a robot capable of interpreting music to create a work of art. It combines the mechatronic field of engineering (building the robot's mechanism, designing the electronics for the sensors and actuators, and developing the programs to control the robot) with the theory of music analysis and interpretation. 

So far, a robot has been designed in the form of an arm that can move in a plane over a canvas. A brush is attached to the end effector. The end effector can rotate to adjust the orientation of the brush and can be raised to control the pressure of the brush on the canvas. Six compartments hold the water and paint (three primary colors, plus black and white), which can be mixed to create an infinite variety of colors. 

This project is broad and covers several aspects of musical interpretation. The first one is to extract some specific parameters of the music (such as BPM, loudness, centroid, genre, and mood) and map them to specific parameters of the robot control (such as speed, color, pressure, and shape). Another aspect is based on the analysis of human music interpretation, where drawings made by humans (mostly children) while listening to music are used to feed a machine learning algorithm to control the robot. Another aspect is the integration of dancers in the music interpretation loop to gesture control the robot. These different aspects will lead to different algorithms to control the robot, leading to different paintings that can then be compared. By comparing the different outputs and processes with those of humans, the aim is to gain a better understanding of human music interpretation. 

History

Results of the ADMM 2023 groups - Robots painting to music
Credit: Ornella Tortorici Pabst

This project started in March 2023 in the context of the ADMM (Applied Design Methodology for Mechatronics) course offered by TUHH. In this course, students learn the methodology of design by carrying out a project with a different theme each year. For that semester, four teams of students each developed a painting robot that interprets music. Following the success of this project, we are continuing to develop it with the help of student theses on a wide range of topics. 

Collaboration with other institutes

This project, initiated by the Haptic Lab (TUHH), integrates other clusters and undertakings. The music analysis and interpretation part will be improved with the support of the ArtSearch Lab, the XR Lab, the Prod Lab and the Inno Lab.  Workshops on music interpretation are organised in schools with the support and pedagogical expertise of colleagues from the HfMT. Public events for workshops and performances will be organised with the support of ligeti's transfer office. 

Description of the robot

So far, a robot has been designed in the form of an arm that can move in a plane over a canvas. A brush is attached to the end effector. The end effector can rotate to adjust the orientation of the brush and can be lifted to control the pressure of the brush on the canvas. Six compartments hold the water and paint (three primary colors, plus black and white), which can be mixed to create an infinite variety of colors. 

Credit and previous works

Oct. - Dec. 2023 Design and implementation of a painting robot capable of authentically imitating human painting for show purposes. Maximilian Finn Lüders Bachelor Thesis
Nov. 2023 - Feb. 2024 Control design and implementation for a painting robot based on music analysis in real time. Jan-Vincent Perlitschke Bachelor Thesis
Apr. - Aug. 2024 Low level control and HMI for a robot interpreting music to paint Irwin Lim Project Work
Mar. - Sept. 2024 Cross-modal generation from music to painting via deep learning Berk Ali Çam Project Work
May. - Nov. 2025 Design of a digital twin for a 4-DOF arm painting robot. Benjamin Krantz Master Thesis
May. - Nov. 2025 Development of a drawing simulator to experiment with different music interpretation parameters Mohammed Mohtesham Ahmed Soherwardi Project Work
Jul. - Oct. 2025 The body as a paintbrush: extending the painting robot through AI-supported body interaction Ivan Iovine Residency

Open topics