Touchscreens are increasingly being used in cars, motorcycles, aircraft, ships, and agricultural machinery to access a wide range of vehicle functions. The primary motivation for incorporating touchscreens in vehicles is that they offer several advantages over physical mechanical controls, including inexpensive to pro- duce, lightweight, low space requirements, design flexibility to handle multiple input/output, quick and easy interface modification, and easy replacement. Touch- screens, on the other hand, lack some features that physical controls have, such as tactile feedback and the same tactile sensations for all controls. The absence of these features on a touchscreen increases visual attentional demands and re- duces driving performance, potentially posing a serious safety risk. We have set a primary goal for this research in order to address these issues: Develop new touchscreen interaction methods to improve driving performance by reducing visual attentional demands. We have set three objectives to achieve the primary goal of this research: (1) Examine the design and use of layout-agnostic stencil overlays for in-vehicle touchscreen; (2) To propose in-vehicle dashboard controls interaction framework; (3) To empirically characterise proprioceptive target acquisition accuracy for in-vehicle touchscreens while driving.

Addressing goal (1). Prior stencil based studies suggested that stencil overlays can reduce the need for visual attention on the touchscreen while driving. However, those stencils were Layout-specific with cuts and holes at the underlying touch- screen controls’ location. As a result, each stencil could only be used with a single underlying interface. Because contemporary in-vehicle touchscreens are almost always multi-functional, with different interface layouts in different parts of the interface, this restriction is unrealistic for in-vehicle touchscreens. To address the limitations of previous stencil-based studies. We aimed to design Layout-agnostic stencils. Layout-agnostic means that one stencil can provide tactile guidance to user interface targets regardless of the underlying interface layout, with the term layout agnostic’ capturing our intention that the stencils should provide tactile guidance to user interface targets regardless of the underlying interface layout. We designed several versions of layout-agnostic stencils iteratively and evaluated them in a simulated driving scenario. Our layout-agnostic stencils failed to reduce visual attentional demands and worsen driving performance, according to the findings.

Addressing goal (2). The failure of objective one prompted us to take a different approach in order to continue working on the research’s main goal. In this regard, we have set a new objective, aiming to yield a new understanding. Our stencils failed despite the iterative design process of layout-agnostic stencils, which was supported by prior studies that showed stencils could reduce visual attentional de- mands. We proposed a “In-vehicle dashboard controls interaction framework” to identify the root causes of layout-agnostic stencils failure. The framework allows for a better understanding of how the driver interacts with the vehicle’s dash- board controls. The framework could be used to create new dashboard interaction techniques as well as evaluate current ones.

Addressing goal (3). We used the proposed framework to evaluate the results of layout-agnostic stencils and discovered three knowledge gaps regarding human- dashboard controls interaction while driving. The first knowledge gap was a lack of understanding of how precisely a human can use proprioception to reach a dash- board control. In this regard, we set another goal and conducted an experimental study to assess human proprioceptive abilities to reach dashboard controls in a simulated driving scenario in terms of distance from the body. We empirically characterise proprioceptive target acquisition accuracy for in-vehicle touchscreens while driving based on experimental results. From various distances, we can now determine how accurately humans can reach a specific location on the touchscreen. We proposed touchscreen control sizes (in cm) based on the characterisation. Ex- isting touchscreen user interfaces could be modified to enable eyes-free proprioceptive target acquisition while driving, which would improve touchscreen interaction safety, based on our recommended touchscreen control sizes.

In conclusion, this thesis makes two minor and one major contribution to the field of in-vehicle touchscreen research. The minor contribution is as follows: (1) Better understanding the use of stencil overlays for in-vehicle touchscreens. The following are the major contributions: (2) We proposed a novel framework and it is the first framework in the vehicle dashboard interaction research domain to the best of our knowledge. The proposed framework provides a better understanding of how drivers interact with dashboard controls in vehicles. (3) We proposed a characterisation of the accuracy of proprioceptive target acquisition for in-vehicle touchscreens while driving.