Augmented reality target finding based on tactile cues

This study is based on a user scenario where augmented reality targets could be found by scanning the environment with a mobile device and getting a tactile feedback exactly in the direction of the target. In order to understand how accurately and quickly the targets can be found, we prepared an experiment setup where a sensor-actuator device consisting of orientation tracking hardware and a tactile actuator were used. The targets with widths 5°, 10°, 15°, 20°, and 25° and various distances between each other were rendered in a 90° -wide space successively, and the task of the test participants was to find them as quickly as possible. The experiment consisted of two conditions: the first one provided tactile feedback only when pointing was on the target and the second one included also another cue indicating the proximity of the target. The average target finding time was 1.8 seconds. The closest targets appeared to be not the easiest to find, which was attributed to the adapted scanning velocity causing the missing the closest targets. We also found that our data did not correlate well with Fitts' model, which may have been caused by the non-normal data distribution. After filtering out 30% of the least representative data items, the correlation reached up to 0.71. Overall, the performance between conditions did not differ from each other significantly. The only significant improvement in the performance offered by the close-to-target cue occurred in the tasks where the targets where the furthest from each other.