The absence of haptic feedback on touchscreens creates a big reliance on visual attention, leading to reduced focus and increased cognitive load, especially in dynamic real-world scenarios like driving or using interactive media. While previous research has extensively explored static pointing tasks, a significant gap exists in understanding interactions with moving targets. This thesis investigates the effect of active lateral force feedback, generated specifically by the Ultraloop haptic device, on user performance in dynamic point-following-tasks.
An experiment with 11 participants was carried out to separate the effect of vision and compare different types of haptic feedback across four conditions: Vision Only, Vision with Active Force, No Vision with Friction Modulation, and No Vision with Active Force. In this task, participants moved their finger on the Ultraloop to track a target that moved along a one-dimensional randomized multi-sine path. Then, their performance for each condition was evaluated by measuring the mean tracking error during their experimental trials. Then, their performance was evaluated with a McRuer crossover model analysis to measure human control behavior during the continuous pointing task by estimating parameters such as crossover frequency, lead and lag for each experimental condition.
The results demonstrate that active lateral force feedback significantly minimizes tracking error. It served as an additional cue in visual conditions, reducing the average tracking error by 24.1%, and acted as an effective sensory substitute in non-visual conditions, drastically outperforming friction modulation feedback by reducing the average tracking error by 34.8%. The McRuer analysis showed that it is possible to identify a human control behavior for visual conditions, but becomes quite complicated for non-visual conditions. Also, participants employed predictive control in visual conditions but reverted to reactive, heavily filtered strategies when relying solely on haptic feedback.
In conclusion, this experiment demonstrated that programmable active lateral force feedback is an effective tool for improving touch-enabled interactions in dynamic tasks. These findings expand the usability of active surface haptics, thereby advancing the development of more intuitive and efficient haptic touch interfaces in the future.