Towards More Transactional Voice Assistants: Investigating the Potential for a Multimodal Voice-Activated Indoor Navigation Assistant for Blind and Sighted Travelers

Title of the Paper

Towards More Transactional Voice Assistants: Investigating the Potential for a Multimodal Voice-Activated Indoor Navigation Assistant for Blind and Sighted Travelers

Paper Information

• Subject Area: Accessibility Technology and Human-Computer Interaction Design
• Keywords: Accessibility, Voice Assistant, Indoor Navigation, Blind and Sighted Travelers, Multimodal Interaction, HCI Design, Air Travel

Research Background and Issues

• What problems or challenges did the authors identify?

  • Voice assistants are typically used for simple tasks (e.g., playing music or checking the weather) but lack designs that support complex task scenarios.
  • Indoor navigation during air travel is highly challenging for both blind and sighted travelers, especially in environments with inconsistent layouts and delayed access to information.
  • Blind individuals, who often prefer technological assistance over human help, are not adequately considered in the design of existing voice assistants.

• Why is this issue important?

  • Indoor navigation during air travel affects not only blind individuals but also sighted travelers, encompassing a broader user base. Extending the application of voice assistants to these scenarios can enhance convenience and reduce reliance on human assistance.
  • Integrating accessibility design into mainstream technology can reduce social stigma and increase adoption rates.

• Research Motivation and Related Work

  • Previous studies on indoor navigation and voice assistants have primarily focused on single disability groups (e.g., blind individuals) or simple voice interaction tasks, without adequately addressing the needs of diverse user groups in complex scenarios.
  • This study aims to expand the functionality of voice assistants by investigating the navigation and information needs and challenges of both blind and sighted travelers during air travel and designing solutions accordingly.

Solution

• What methods or solutions did the authors propose?

  • The authors propose extending mainstream voice assistants into a voice-activated, multimodal air travel navigation assistant. This would cover tasks ranging from trip planning to airport navigation and handling complex scenarios such as flight connections and transportation arrangements.
  • Design recommendations include voice commands, real-time notifications, and multimodal feedback (e.g., vibrations, auditory cues, and customized voice outputs).

• What is innovative about this solution?

  • Transition from simple to complex tasks: Expanding the interaction model of voice assistants from "human-human" dialogue to a transactional model.
  • Support for multimodal interaction: Integrating voice, visual, and tactile feedback to optimize interaction experiences across different scenarios.
  • User customization and contextual adaptation: Allowing adjustments to feedback frequency, speech rate, interaction modes, etc., to meet individual needs.

• What are the implementation steps? What key technologies were used?

  • Research Design: Conducted mixed-group focus interviews involving both blind and sighted participants.
  • Scenario Analysis: Extracted common challenges and needs from participants' feedback on airport navigation scenarios.
  • Interaction Design: Designed voice interaction logic supporting various operational modes (e.g., planning, navigation, exploration) based on user suggestions.

Research Outcomes

• What specific outcomes were achieved?

  • Identified common challenges faced by blind and sighted travelers during airport navigation (e.g., inconsistent layouts, delayed access to information), highlighting overlapping technical needs between the two groups.
  • Listed user-recommended voice assistant features (e.g., tracking trip progress, navigation paths, providing transportation options) and specific interaction examples.
  • Proposed a voice assistant design framework based on a transactional model and multimodal interaction.

• How does it compare to existing solutions?

  • Extends the application scope of existing voice assistants from supporting single tasks to complex contextual tasks.
  • Offers a more universal and personalized design to meet the needs of diverse users, rather than being limited to a single group (e.g., visually impaired individuals).
  • Integrates indoor navigation, dynamic information needs, and voice-based interaction into a single system, eliminating the hassle of switching between different systems.

• What were the experimental or evaluation results?

  • Focus group interviews revealed significant shared challenges between blind and sighted participants, and both groups recognized the potential of voice assistants in indoor navigation environments.
  • Users expressed specific expectations for how the system could support their planning, navigation, and operations, such as precise localization, multi-step planning, and customized route prompts.

• Limitations and Future Directions

  • Limitations:
    • Small sample size of only 18 participants, which may not fully represent a broader user base.
    • Did not explore the needs of group travel or users with other types of physical disabilities in depth.
  • Future Directions:
    • Expand the user base to include individuals with disabilities, older adults, and those with multiple impairments.
    • Conduct iterative online design processes to further validate and refine interaction designs.
    • Explore the integration capabilities of voice assistants with broader services or technologies.