Suh, Youngbo (2018-08). Supporting Visual Attention Management of Drivers Interacting with In-Vehicle Touchscreens. Doctoral Dissertation. Thesis uri icon

abstract

  • Touchscreens are widely used in interface designs, with the benefits of ease of use and flexibility in interaction design. However, visual attention is essential to touchscreen interaction, which results in frequent reorientation of attention in multitask contexts and thus leads to task interferences. This visually-demanding nature of touchscreens can expose operators to safety risks in certain domains, such as in driving. In this regard, earlier studies have attempted to address visual distraction potential related to in-vehicle touchscreens. Nevertheless, road traffic statistics show that in-vehicle touchscreens are still contributing to a considerable number of crashes each year. This dissertation aims to investigate and design means to offload visual demands of touchscreen interaction in complex, data-rich environments, focusing on driving as the application domain. The research tested the configurations of an in-vehicle touchscreen, concerning feedback modality and control size. Vertical locations of this secondary interface were also examined in terms of distraction effects. Preliminary design efforts evaluated the practicality of a symbolic encoding method using vibrotactile beats. A novel interface design, Nonvisual Aids for Touchscreen Experience (NATE), primarily employs distributed tactile cues that are proximal to operators but physically separated from a touchscreen device. This method involves presenting vibratory cues to the back of operators, and each cue is spatially mapped to each touchscreen control. NATE allows operators to bind those multiple stimuli to perform touchscreen interaction via nonvisual channels. An evaluation study was conducted in a simulated environment, comparing measures in driving performance, glance behavior, and perceived workload between an ordinary touchscreen and NATE-based touchscreen applications. The driving scenarios were based on the Lane Change Test procedure established by the ISO standard. The findings indicate that the application of NATE to in-vehicle touchscreen designs significantly reduced visual distraction and improved performance in vehicle operation. However, performance measures on a touchscreen task suggest limitations. This body of research provides inputs to the design community of in-vehicle touchscreens, focusing on offloading visual demands of drivers. Theoretical contributions include a deeper understanding of glance behavior, which can inform models of human behavior in multitask contexts.
  • Touchscreens are widely used in interface designs, with the benefits of ease of use and flexibility in interaction design. However, visual attention is essential to touchscreen interaction, which results in frequent reorientation of attention in multitask contexts and thus leads to task interferences. This visually-demanding nature of touchscreens can expose operators to safety risks in certain domains, such as in driving. In this regard, earlier studies have attempted to address visual distraction potential related to in-vehicle touchscreens. Nevertheless, road traffic statistics show that in-vehicle touchscreens are still contributing to a considerable number of crashes each year.
    This dissertation aims to investigate and design means to offload visual demands of touchscreen interaction in complex, data-rich environments, focusing on driving as the application domain. The research tested the configurations of an in-vehicle touchscreen, concerning feedback modality and control size. Vertical locations of this secondary interface were also examined in terms of distraction effects. Preliminary design efforts evaluated the practicality of a symbolic encoding method using vibrotactile beats.
    A novel interface design, Nonvisual Aids for Touchscreen Experience (NATE), primarily employs distributed tactile cues that are proximal to operators but physically separated from a touchscreen device. This method involves presenting vibratory cues to the back of operators, and each cue is spatially mapped to each touchscreen control. NATE allows operators to bind those multiple stimuli to perform touchscreen interaction via nonvisual channels. An evaluation study was conducted in a simulated environment, comparing measures in driving performance, glance behavior, and perceived workload between an ordinary touchscreen and NATE-based touchscreen applications. The driving scenarios were based on the Lane Change Test procedure established by the ISO standard.
    The findings indicate that the application of NATE to in-vehicle touchscreen designs significantly reduced visual distraction and improved performance in vehicle operation. However, performance measures on a touchscreen task suggest limitations. This body of research provides inputs to the design community of in-vehicle touchscreens, focusing on offloading visual demands of drivers. Theoretical contributions include a deeper understanding of glance behavior, which can inform models of human behavior in multitask contexts.

publication date

  • August 2018