Case Study: Georgia Tech builds humanlike robot trust and expressive dance performances with MATLAB and Simulink from MathWorks

A Forest of Robots Builds Trust Through Dance

The Center for Music Technology at Georgia Tech, a research group at the university, sought to build public trust in robots by making their movements and sounds appear more emotional and humanlike. They faced the challenge of programming robotic systems to express nuanced social cues through fluid motion and audio, requiring sophisticated control and simulation to avoid unnatural, jerky movements that could damage hardware and fail to connect with people.

Using MathWorks products MATLAB and Simulink, the team developed the system for their "FOREST" project. The solution allowed them to program, simulate, and visualize complex robotic movements, generate smooth motion paths from motion-capture data, and train neural networks to create emotional audio and movement styles. The result was a successful live and videotaped performance where humans and robots danced together, significantly increasing the dancers' comfort and trust in their mechanical partners and demonstrating a viable path toward more natural human-robot collaboration.

Georgia Tech

Gil Weinberg

Founding Director