Haptixel: Encoding Data through Cutaneous Force-based Encountered-Type Fingertip Haptics

Paper Title

Haptixel: Encoding Data through Cutaneous Force-based Encountered-Type Fingertip Haptics

Publication Info

• Topic area: Use of cutaneous fingertip haptics for data visualization.
• Keywords: Haptics, data visualization, cutaneous feedback, force-feedback, wearable devices, fingertip haptics, multimodal interaction, tactile interfaces, data physicalization, VR/AR.

Background and Problem

• Problem / challenge: Existing data visualization methods relying on non-visual cues are often cumbersome, require large-scale artifacts, or depend on external devices, limiting their adaptability to dynamic or complex datasets. Cutaneous force-feedback as a modality for visualization remains underexplored.
• Significance: Enabling users to perceive and interact with data through lightweight, wearable haptic devices can enhance multimodal visualization, particularly in scenarios where visual attention is limited or additional dimensions need to be conveyed.
• Motivation and related work: Prior work has explored multimodal visualization using sound, tactile cues, and vibrotactile feedback, but these approaches often rely on grounded devices or external artifacts. Wearable haptic devices for fingertip cutaneous feedback offer a more seamless and realistic interaction modality, but their potential for quantitative data visualization tasks has not been fully investigated.

Solution

• Proposed approach: Haptixel, a lightweight, open-source, DIY wearable device providing cutaneous force-feedback to the fingertip pulp, enabling users to perceive data values through skin deformation.
• Novelty:
  1. Development of a compact, unobtrusive fingertip haptic device that supports dynamic interaction and integrates with VR/AR systems.
  2. Empirical quantification of cutaneous force-discrimination thresholds (Just Noticeable Difference, JND) and the impact of finger support material on perception.
  3. Demonstration of Haptixel's ability to enable accurate data retrieval and pattern reconstruction using force cues alone.
• Procedure and key techniques:
  • Haptixel uses a crank-rod mechanism to provide force-feedback through fingertip skin deformation.
  • The device is servoed in force and controlled via Unity3D, with an average latency of 1.4 seconds and an accuracy below 0.07 N.
  • Interaction framework includes visual support, force levels (low to high), and contact types (discrete, continuous, or varying).
  • Preliminary study quantified JND values for different finger support materials, followed by a user experiment where participants reconstructed patterns using force cues.

Results

• Concrete findings:
  • JND for force discrimination was 0.6 N with a flexible finger support, allowing at least four distinguishable force levels.
  • Participants achieved an average absolute error of 0.34 N in force retrieval tasks and accurately reconstructed patterns using force cues alone.
  • Relative force differences were consistently perceived, enabling clustering and comparison tasks.
• Advantage over baselines:
  • Demonstrated finer discrimination levels (four force levels) compared to vibrotactile feedback (three levels) and comparable to kinesthetic feedback while being more lightweight and unobtrusive.
  • Effective for both absolute and relative data retrieval tasks without visual encoding.
• Experiments / evaluation:
  • Preliminary study (n=8) evaluated JND for rigid (PLA) and flexible (TPU) finger supports.
  • Main user experiment (n=16) involved reconstructing 3D patterns in a VR environment using force cues, with tasks including discovery, retrieval, and clustering.
• Limitations and future work:
  • Current granularity limited to one fingertip-scale pixel at a time.
  • Maximum force (3 N with flexible support) restricts the number of discriminable levels.
  • Latency could be reduced for finer-grained visualizations.
  • Future work includes integrating additional cutaneous cues (e.g., vibrations), exploring alternative actuation mechanisms, and assessing cognitive load in multimodal tasks.

Summary

Haptixel is a lightweight, open-source fingertip haptic device designed to encode data through cutaneous force-feedback, enabling users to perceive and interact with data dynamically. Empirical studies demonstrated that Haptixel supports accurate discrimination of at least four force levels, enabling tasks such as data retrieval, clustering, and pattern reconstruction without visual encoding. The device offers a promising alternative to vibrotactile and kinesthetic feedback for visualization tasks, with applications in spatial and temporal data exploration. Future work will address granularity, force range, and integration with additional haptic cues to enhance its applicability.