Head-mounted Sensory Augmentation System for Navigation in Low Visibility Environments

Sensory augmentation can be used to assist in some tasks where sensory information is limited or sparse. This thesis focuses on the design and investigation of a head-mounted vibrotactile sensory augmentation interface to assist navigation in low visibility environments such as firefighters’ navigation or travel aids for visually impaired people. A novel head-mounted vibrotactile interface comprising a 1-by-7 vibrotactile display worn on the forehead is developed. A series of psychophysical studies is carried out with this display to (1) determine the vibrotactile absolute threshold, (2) investigate the accuracy of vibrotactile localization, and (3) evaluate the funneling illusion and apparent motion as sensory phenomena that could be used to communicate navigation signals. The results of these studies provide guidelines for the design of head-mounted interfaces. A 2nd generation head-mounted sensory augmentation interface called the Mark-II Tactile Helmet is developed for the application of firefighters’ navigation. It consists of a ring of ultrasound sensors mounted to the outside of a helmet, a microcontroller, two batteries and a refined vibrotactile display composed of seven vibration motors based on the results of the aforementioned psychophysical studies. A ‘tactile language’, that is, a set of distinguishable vibrotactile patterns, is developed for communicating navigation commands to the Mark-II Tactile Helmet. Four possible combinations of two command presentation modes (continuous, discrete) and two command types (recurring, single) are evaluated for their effectiveness in guiding users along a virtual wall in a structured environment. Continuous and discrete presentation modes use spatiotemporal patterns that induce the experience of apparent movement and discrete movement on the forehead, respectively. The recurring command type presents the tactile command repeatedly with an interval between patterns of 500 ms while the single command type presents the tactile command just once when there is a change in the command. The effectiveness of this tactile language is evaluated according to the objective measures of the users’ walking speed and the smoothness of their trajectory parallel to the virtual wall and subjective measures of utility and comfort employing Likert-type rating scales. The Recurring Continuous (RC) commands that exploit the phenomena of apparent motion are most effective in generating efficient routes and fast travel, and are most preferred. Finally, the optimal tactile language (RC) is compared with audio guidance using verbal instructions to investigate effectiveness in delivering navigation commands. The results show that haptic guidance leads to better performance as well as lower cognitive workload compared to auditory feedback. This research demonstrates that a head-mounted sensory augmentation interface can enhance spatial awareness in low visibility environments and could help firefighters’ navigation by providing them with supplementary sensory information