ProxSituated Visualization: An Extended Model of Situated Visualization using Proxies for Physical Referents

Title of the Paper

ProxSituated Visualization: An Extended Model of Situated Visualization using Proxies for Physical Referents

Paper Information

• Subject Area: Data Visualization, Immersive Analytics, Augmented Reality, and Virtual Reality
• Keywords: Situated Visualization, ProxSituated Visualization, Immersive Analytics, Augmented Reality, Data Physicalization, Digital Twins, Embedded Representation, Spatial Indirection, Temporal Indirection, Representation Scale

Research Background and Problem

Identified Problems or Challenges

1. Limitations of Current Models: Existing "situated visualization" models primarily focus on the relationship between data and its directly associated physical objects and environments (physical referents). However, these models lack the descriptive capability when interactions are based on proxy representations of these physical referents.
2. Diverse Needs: In many real-world scenarios, users may not have direct access to physical referents (e.g., remote analysts only interacting with a digital 3D model of a device).
3. Lack of Concepts and Terminology: In the context of using proxies, existing literature lacks clearly defined terminology, making it difficult to classify and systematically describe such work.
4. Untapped Design Space Potential: The development of hybrid approaches combining proxy representations and data representations, including visualizations of historical data and simulations of predictive data, reveals unexplored design spaces.

Significance

• Proxy representations alleviate the strict spatial and temporal constraints of situated visualization, enabling visualization in a broader range of contexts.
• Systematic description and evaluation of the current state, as well as the ability to generate new design scenarios, are crucial for advancing research, yet a comprehensive framework is still lacking.

Research Motivation and Related Work

1. Origins and Progress in Related Work:

   • Early work (e.g., White et al.'s SiteLens system in 2006) demonstrated the potential of situated visualization by integrating data visualization into physical environments.
   • The model proposed by Willett et al. in 2016 further developed the foundation of situated visualization but still cannot describe proxy representations.

2. Technological Drivers: Rapid advancements in augmented reality, virtual reality, and 3D printing technologies provide cost-effective means for implementing proxy-based situated visualizations.

Proposed Solution

Innovative Approach

• ProxSituated Visualization Model: Extends the existing situated visualization model by defining ProxSituated Visualization as a visualization technique encompassing both Immediate Situated Visualization (direct situated visualization) and ProxSituated Visualization (proxy-based situated visualization).
• Model Structure:
  1. Introduction of new entities: physical environment, physical referent proxy, and environment proxy.
  2. Description of relationships between entities: including spatial indirection, temporal indirection, representation scale, and representation form.

Implementation Steps and Key Techniques

1. Modeling Process:
   • Incorporate core entities from existing models (e.g., data, physical referents, visualization pipeline).
   • Extend the model to include proxy representations (e.g., physical referent proxies and environment proxies) and their relationships with data and the environment.
2. Analysis of Relationship Characteristics:
   • Covers spatial and temporal indirection, contrasting historical, real-time, and predictive data proxy representations.
   • Quantifies semantic and geometric scaling between data and physical referents.
3. Design Space Generation:
   • Constructs a multidimensional design space to describe and evaluate existing visualization work and generate new situated design scenarios.

Research Outcomes

Specific Achievements

1. Proposed a new model and standardized terminology (e.g., "ProxSituated Visualization") to create a more refined design framework.
2. Analyzed existing literature using the design space, systematically summarizing major trends in proxy-based situated visualization design, such as:
   • The prevalent use of historical proxy representations.
   • Optimized design cases displaying multivariate data on embedded dashboards.
   • A significant increase in the use of digital proxies to describe physical objects.
3. Demonstrated the differences between proxy-based and direct situated visualizations and the potential of their combined use.

Experimental and Evaluation Results

1. By reviewing 43 papers related to proxy-based situated visualization, the authors matched existing work with the model's spatial indirection, temporal indirection, representation form, and scale.
   • Examples include use cases for real-time, predictive, and historical proxies.
2. Highlighted the advantages of proxy-based situated visualization:
   • Suitable for scenarios where physical referents are remote.
   • Provides flexible solutions for dynamic referents (e.g., data streams of athletes) and static referents (e.g., architectural design models).

Comparison with Existing Models

• Compared to Willett et al.'s situated model, ProxSituated Visualization is more expressive, encompassing design scenarios under proxy representation conditions.
• Explored various untapped areas, such as integrating physicalized datasets into proxy environments.

Limitations and Future Directions

1. Limitations:
   • Cost and time investment pose significant practical challenges; future research needs to balance the cost-effectiveness of proxy representations and on-site visualizations.
   • The impact of semantic scaling and proxy design on user experience has not been fully quantified.
2. Future Directions:
   • Combine direct and proxy-based situated visualizations, such as supporting real-time collaboration in single-user or multi-user scenarios.
   • Develop real-time dynamic proxies to expand support for data stream scenarios.
   • Explore proxy representations of non-visual environmental factors (e.g., sound, smell).
   • Integrate into digital twins, smart environments, and immersive physicalization domains to better align with interdisciplinary research directions.

Conclusion

By extending the existing situated visualization model, the authors demonstrate how proxy-based representations enrich the design space and meet the demands of data analysis in complex scenarios. ProxSituated Visualization provides a theoretical framework and an inspiring design tool for both academic and practical applications, paving the way for further exploration of visualization technologies.