Tapsonic: One dimensional finger mounted line chat reader

 

This project was inspired by the insufficient research on inclusive design of tactile charts for visually imparied group. Based on our presvious codesign workshop and literature review, it was found that the perceputaion of 2D information was usually inaccurate and cognitive demanding. To reduce workload, I propose a ‘dimension reduction strategy’ to simplify a 2D patterns into bi-directional haptics and unidimension linear motion. Two natural guestures were thereby introduced to further enpower the accessbility.

Please see our co-design session to understand our co-design procedure.
_____________________________________________________

Supervisor: Albert Higgins, Catherine Holloway
Extended abstract accepted by ASSETS 2020 Student Research Competition

assets2020logo.png
 
 
 

Interaction —

The interface sim­plifies two-dimensional line graph into a unidimensional movement (horizontal) and bidirectional haptics (which substitutes vertical dimension). Users can adopt swipe and pinch gestures to read and compare single or multiple lines together.

  • By swiping horizontally, the interface will detect selected data and clarify its value by voiceover.

  • By swiping vertically, users can change data series. The top and bottom vibrations are induced separately by positive and negative value change.

  • To prevent voiceover overlap, auditory clarification is only activated when sufficient idle time is satisfied.

  • To find inter­sections of two data series, users can adopt a pinch gesture to compare value difference of two series on the same position of the horizontal axis. Positive and negative value difference will trigger top and bottom linear actuator accordingly.

 
Swipe horizontally to read data values (no need to follow the physical shapes). Auditory and haptic feedback would be provided to assist interpret the graph

Swipe horizontally to read data values (no need to follow the physical shapes). Auditory and haptic feedback would be provided to assist interpret the graph

 

Methodology —

Four participants (1 female), aged between 22 to 27 years old (mean=25.2, SD=1.7) were voluntarily recruited. All candidates had congenital visual impairment and without other disabilities. A fixed design was used to evaluate the performance and gather user feedbacks of the proposed system.

The performance was evaluated by finish time, accuracy rate and cognitive workload. All participants conducted experiments in the following order:

  • A brief introduction of the system and three minutes of self-exploration of the system.

  • An experiment in which they finish given tasks in order, time and accuracy are recorded.

  • Filling in accessible version of NASA task load index (TLX) questionnaire (10-point scale) and system usability scale survey (5-point scale),

  • Attending a followed up by a semi-structured interview.

 

Conclusion —

The initial trials were extremely positive, with the results showed that partic­ipants were able to read graphs in an accuracy up to 85% overall. The system was reported intuitive to use, with great reduction of mental workload compared with conventional tactile graphs. This feature is highlighted especially in exploring in­ dividual data series with swipe gestures. Future research is needed to examine the system functionality in wider scenarios. I hope the design could provide some insights into the design of accessible chart reader via dimensionality reduction approach.

 
 

Please refer to our paper for more information:
https://dl.acm.org/doi/10.1145/3373625.3417075