DataLev: Mid-air Data Physicalisation Using Acoustic Levitation

Document Title

DataLev: Mid-air Data Physicalisation Using Acoustic Levitation

Document Information

• Subject Area: Data Physicalisation, Human-Computer Interaction, Acoustic Levitation Technology
• Keywords: Data Physicalisation, Acoustic Levitation, Interactive Systems, Human-Computer Interaction, Multimodal Support, Mixed Reality, Animation

Research Background and Problems

• Identified Problems or Challenges:
  1. Current data physicalisation techniques struggle to simultaneously achieve reconfigurability and material diversity.
  2. Existing methods for data physicalisation require trade-offs between flexibility and material freedom when choosing static or dynamic implementations.
  3. Most dynamic physicalisation methods, such as shape-changing devices or swarm robots, only support limited 2.5D structures, making true 3D manipulation difficult.
• Importance of the Problems: Data physicalisation provides tools for deeper understanding of data and multisensory experiences, but technical limitations in current methods hinder advancements in interactivity and expressiveness.
• Research Motivation and Related Work:
  • The need for dynamic physicalisation and material flexibility drives exploration of alternative methods, such as acoustic levitation (acoustophoresis).
  • Acoustic levitation technology enables manipulation of diverse materials and achieves 3D dynamic reconfiguration through sound waves, opening new possibilities for data physicalisation.

Solution

• Proposed Method or Solution: The authors designed and implemented a platform called DataLev, which leverages acoustic levitation technology to achieve hollow reconfigurability, material diversity, and multimodal data physicalisation.
• Innovative Contributions:
  1. Proposed a design space for data physicalisation encompassing five dimensions: embodiment, material, multimodal support, mixed reality components, and animation.
  2. Provided hardware and software solutions, including projection mapping, ASKA3D plates, and path planning techniques, to support more dynamic physicalisation scenarios.
• Implementation Steps:
  1. Used open ultrasonic arrays (16×16 arrays) to create various levitation configurations (e.g., top-bottom, V-shaped, or single-sided setups).
  2. Employed efficient path planning techniques to achieve dynamic animations, ensuring safe movement and state transitions of multiple objects in 3D space.
  3. Applied mixed reality components (e.g., ASKA3D plates) to enhance the integration of digital and physical representations.
  4. Supported diverse materials (including solid particles, liquids, and food) and utilized self-assembly and automated injection techniques for initial material positioning.

Research Outcomes

• Specific Results:
  1. Proposed an actionable design space and demonstrated how the DataLev platform supports dynamic and material-rich physicalisation through eight examples.
  2. Quantified the platform's performance through technical evaluations, including reconfigurability and material support capabilities.
• Advantages Over Existing Solutions:
  1. Supports multimodal interaction, combining visual, tactile, auditory, and even gustatory experiences.
  2. Successfully decouples material selection from reconfigurability.
  3. Enables full 3D animations and dynamic data representations, offering greater flexibility compared to previous 2.5D or static physical models.
• Experimental or Evaluation Results:
  • Tests conducted with multiple data points and different animation dimensions (2D/3D) revealed that balancing animation speed with the number of data points is crucial for maintaining levitation stability.
  • Levitation experiments with various liquid materials (e.g., water, milk, isopropanol) showed that material density and surface tension significantly impact performance.
• Limitations and Future Directions:
  1. The current platform faces limitations in animation speed and the number of data points, requiring further optimization of hardware and algorithms.
  2. Tactile experiences in multimodal interactions remain inferior to direct touch.
  3. The platform lacks plugin support for external visualization tools (e.g., Excel or D3.js), suggesting future exploration of a more comprehensive end-to-end data physicalisation pipeline.
  4. Further research is needed to extract clearer design principles to guide future studies on acoustic levitation-based data visualisation.

This document presents an innovative platform in the field of data physicalisation, offering significant insights and potential applications in both technical implementation and human-computer interaction design.