Daniel PahrORCID iD
From Interactions to Integrated Actions: Exploring Active Perception and Inter-Modality in Data Physicalization
Supervisor: Renata Georgia RaidouORCID iD
Duration: 2021 — 2024
[thesis]

Information

Abstract

The growing field of data physicalization holds significant potential for integrating user actionsdirectly into the sense making process through physical artifacts. Two promising factors for physical, as opposed to virtual representations, are physical interaction and multimodal perception. Unmediated interaction in the physical space allows users to manipulate and explore dataphysicalizations in a natural way, harnessing a user’s actions to encode and decode information ina different way than purely virtual representations. In this dissertation, I explore the incorporation of user action as a means of manipulation and perception into data physicalizations, moving from representations where perception only happens after physical interactions, to representations where physical interactions directly stimulate the user’s perception. I investigate four distinct types of user interactions with data physicalizations and show how each of them can support human perception in different ways. Firstly, I show how a modular 3D representation of dynamic data can leverage physical embodiment using natural spatial perception.I demonstrate this by creating a simple interactive physical representation of a space-time-cubemetaphor and investigating it in a case study with a domain expert. Secondly, I investigate the influence of construction — an intuitively physical interaction in the physical space — of apre-defined physical representation on human perception. I show this by designing a networkdata physicalization toolkit and conducting a between-subject study, comparing different ways to instruct a user during construction. Thirdly, I introduce tactile perception of the elastic properties of an object in a multi-modal representation of volume data. I showcase this at the hands of a fabrication pipeline that creates elastic artifacts from volume data using consumer-level 3D printing and validate the method through computational, mechanical, and perceptualstudies. Finally, I explore the benefits of manually operating a physical representation of adynamic process, leveraging the tactile feedback to the user for information encoding. By means of a between-subject user study, I show that integrating a user’s actions into a representation significantly increases engagement.Overall, the results show that even a simple physicalization can highlight the perceptual benefits of physically encoding data by ways of natural perception. Abstract representations have to be learned by users but can be supported by physical interactions, while embodied metaphors profit from direct interactivity if the stimulus fits the sensory capabilities.

Additional Files and Images

thesis
thesis

Weblinks

BibTeX

@phdthesis{Pahr_PhD,
  title =      "From Interactions to Integrated Actions: Exploring Active
               Perception and Inter-Modality in Data Physicalization",
  author =     "Daniel Pahr",
  year =       "2025",
  abstract =   "The growing field of data physicalization holds significant
               potential for integrating user actionsdirectly into the
               sense making process through physical artifacts. Two
               promising factors for physical, as opposed to virtual
               representations, are physical interaction and multimodal
               perception. Unmediated interaction in the physical space
               allows users to manipulate and explore dataphysicalizations
               in a natural way, harnessing a user’s actions to encode
               and decode information ina different way than purely virtual
               representations. In this dissertation, I explore the
               incorporation of user action as a means of manipulation and
               perception into data physicalizations, moving from
               representations where perception only happens after physical
               interactions, to representations where physical interactions
               directly stimulate the user’s perception. I investigate
               four distinct types of user interactions with data
               physicalizations and show how each of them can support human
               perception in different ways. Firstly, I show how a modular
               3D representation of dynamic data can leverage physical
               embodiment using natural spatial perception.I demonstrate
               this by creating a simple interactive physical
               representation of a space-time-cubemetaphor and
               investigating it in a case study with a domain expert.
               Secondly, I investigate the influence of construction — an
               intuitively physical interaction in the physical space —
               of apre-defined physical representation on human perception.
               I show this by designing a networkdata physicalization
               toolkit and conducting a between-subject study, comparing
               different ways to instruct a user during construction.
               Thirdly, I introduce tactile perception of the elastic
               properties of an object in a multi-modal representation of
               volume data. I showcase this at the hands of a fabrication
               pipeline that creates elastic artifacts from volume data
               using consumer-level 3D printing and validate the method
               through computational, mechanical, and perceptualstudies.
               Finally, I explore the benefits of manually operating a
               physical representation of adynamic process, leveraging the
               tactile feedback to the user for information encoding. By
               means of a between-subject user study, I show that
               integrating a user’s actions into a representation
               significantly increases engagement.Overall, the results show
               that even a simple physicalization can highlight the
               perceptual benefits of physically encoding data by ways of
               natural perception. Abstract representations have to be
               learned by users but can be supported by physical
               interactions, while embodied metaphors profit from direct
               interactivity if the stimulus fits the sensory capabilities.",
  pages =      "124",
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Research Unit of Computer Graphics, Institute of Visual
               Computing and Human-Centered Technology, Faculty of
               Informatics, TU Wien ",
  keywords =   "Data physicalization, Interaction, Embodiment",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2025/Pahr_PhD/",
}