Collider-based movement detection and control of wearable soft robots for visually augmenting dance performance

Area of Science:

• Robotics and Human-Computer Interaction
• Performance Art Technology
• Biomechanical Motion Analysis

Background:

• Improvisational dance presents unique motion-tracking challenges due to unpredictable, non-repetitive movements.
• Existing methods like optical tracking (limb occlusion) and inertial measurement units (complex algorithms, fixed thresholds) have limitations.
• Machine learning is often unsuitable for improvisational dance due to limited unique training data.

Purpose of the Study:

• To develop an adaptable and robust motion-tracking algorithm for improvisational dance.
• To enable real-time control of wearable soft robotic actuators based on dancer movements.
• To enhance the expressive potential of dance through technology integration.

Main Methods:

• Introduction of a collider-based movement detection algorithm, modeling colliders as virtual mass-spring-damper systems.
• Defining colliders in limb-specific planes to capture dynamics in their relative frame, avoiding IMU drift.
• Implementing a simplified detection scheme using individual dynamic system response variables and a hashing method for high-speed processing.

Main Results:

• The collider-based algorithm effectively detects complex, improvisational dance movements.
• Demonstrated control of wearable, origami-based soft actuators (changing size and lighting) in response to detected movements.
• Experimental validation of the algorithm's robustness and speed for real-time dance applications.

Conclusions:

• The proposed algorithm offers a robust solution for tracking arbitrary dance motions, overcoming limitations of conventional methods.
• This technology allows dancers to trigger stage events organically, creating novel aesthetic experiences.
• The research highlights the potential of integrating advanced robotics and sensing with spontaneous artistic expression to augment performance art.