Crownboard: A One-Finger Crown-Based Smartwatch Keyboard for Users with Limited Dexterity

Document Title

Crownboard: A One-Finger Crown-Based Smartwatch Keyboard for Users with Limited Dexterity

Document Information

• Subject Area: Assistive technology for wearable devices, text entry, and human-computer interaction
• Keywords: smartwatch, wristwatch, crown, virtual keyboard, text entry, accessibility, motor impairments

Research Background and Problem

• What problems or challenges did the authors identify?

  • Existing smartwatch text entry technologies are not user-friendly or suitable for individuals with motor impairments (e.g., tremors, limited fine motor skills, hand shakiness).
  • Smartwatch screens are small, with narrow interfaces, and current methods often require precise target selection or complex gesture operations, making them unsuitable for users with limited mobility.
  • Voice input is not widely used in public settings due to privacy and security concerns, as well as interference from environmental noise.

• Why is this problem important?

  • Improving smartwatch interaction design for individuals with motor impairments can enhance their quality of life and independence, enabling them to perform daily tasks (e.g., sending messages, searching for information) more conveniently.
  • While smartwatches are highly portable, the design of current input methods neglects the needs of specific user groups.

• Research Motivation and Related Work

  • Research on assistive technologies for wearable devices remains limited, particularly in designing for individuals with motor impairments.
  • Several studies have evaluated gesture recognition, rotary-based, or miniaturized keyboard input methods for general users but have not systematically tested their applicability for users with motor impairments.

Solution

• What methods or solutions did the authors propose?

  • The authors proposed Crownboard, a novel text entry technique based on the rotational and press operations of a smartwatch crown.
  • The system uses a circular keyboard layout, dividing the 26 English letters into 8 sectors based on frequency grouping. Users rotate the smartwatch crown to select a sector and press the crown to confirm input.

• What are the innovative aspects of this solution?

  • Physical Feedback: The physical crown on the smartwatch provides tactile feedback, improving hand stability.
  • Streamlined Operations: Reduces the need for direct touchscreen interaction, allowing users to complete most input tasks through single-finger rotation and pressing.
  • Optimized Layout: Letter grouping is optimized based on bigram frequency to reduce input ambiguity.
  • Multiple Modes: Supports both manual rotation and automatic scanning modes to accommodate different user needs.

• What are the implementation steps and key technologies used?

  • Keyboard Design: All possible letter groupings were evaluated to select a layout with minimal conflicting bigrams.
  • Scanning Optimization: In automatic scanning mode, two methods are provided: "clockwise scanning" and "shortest path scanning," with the latter dynamically adjusting scanning direction to accelerate input.
  • Error Correction and Suggestions: A statistical decoder and binary prefix tree (Trie) are used for multi-word prediction and auto-completion.
  • OOV Support: A "double-tap to select letters within a group" mechanism addresses the input of out-of-vocabulary (OOV) words.

Research Outcomes

• What specific results were achieved?

  • User experiments demonstrated that Crownboard is more user-friendly for individuals with motor impairments. All participants with impairments were able to complete tasks using Crownboard, whereas 70% of them were unable to operate the default QWERTY keyboard.
  • In representative user tests, the typing speed in automatic scanning mode was approximately 2.1 words per minute (wpm), with an error rate below 1%.
  • Compared to the default keyboard, Crownboard (manual/automatic versions) improved input speed by 9% and 23%, respectively.

• What advantages does it have over existing solutions?

  • Compared to touch-based or gesture-based input methods, Crownboard significantly improves operational fluency and accuracy, particularly for users with motor impairments.
  • The tactile feedback provided by the crown enhances operational stability, reducing errors caused by involuntary tremors or slips.

• What were the experimental or evaluation results?

  • Multi-phase user experiments were conducted, comparing manual and automatic scanning, as well as clockwise and shortest path scanning, to evaluate system performance.
  • Participants' behavioral preferences and sector selection strategies were observed during the experiments. Some users preferred manual rotation strategies, while others favored the automatic mode to reduce physical effort.

• Limitations and Future Directions

  • Limitations:
    • The current prototype does not support uppercase letters, numbers, or special characters, requiring further optimization to accommodate more comprehensive text entry scenarios.
    • Due to the constraints of the experimental design, the method's long-term learning curve may not be fully reflected.
    • While the shortest path scanning method is theoretically more efficient, its higher learning curve limited its full potential during practical use.
  • Future Directions:
    • Conduct longitudinal studies to assess the impact of user familiarity on system performance.
    • Apply more advanced machine learning models (e.g., LSTMs, Transformers) to enhance the decoder's predictive capabilities.
    • Improve the shortest path scanning algorithm and user interface to better support complex text entry tasks.