Navigating through buildings with impaired vision: challenges and solutions.

Abstract

Navigation is the ability to plan and follow routes between locations, often with an internal or external map of the environment. Vision is an important way to access environmental information for navigation. Consequently, independent navigation is a significant challenge for individuals with visual impairment. This thesis describes three studies that investigate how real or simulated visual impairment affects the ability to navigate inside buildings. Furthermore, these experiments explore methods for compensating for the loss of visual information, either by using other senses or by using assistive technology. The visual information in an environment that is useful for navigation can be categorized as two types: geometric (visual information conveying layout geometry such as hallways and intersections), and non-geometric (features other than geometry such as lighting, texture, and object landmarks). The first experiment (Chapter 2) tested the effects of visual impairment and age on the use of these two types of visual information for navigation. In the second experiment (Chapter 3), visually-impaired individuals were tested on their ability to follow verbal route instructions provided by an indoor navigation technology. The instructions described distances in one of three ways: feet, number of steps, and travel time in seconds. This study compared route-following performance with the three distance modes. The third experiment (Chapter 4) investigated the integration of visual and walking information for localization in a hallway. We predicted that humans integrate information: 1) only when they perceive themselves to be near a landmark after walking (congruency), and 2) by weighing each information source according to its reliability. Normally-sighted participants judged their location in a hallway after viewing a target and then walking blindfolded to either the visual target or to a slightly different location. Participants viewed targets in two conditions that manipulated visual reliability: normal viewing and blurry viewing. This experiment tested and confirmed a statistical model of human perception in a novel domain. Together, these three studies enhance our understanding of the effects of visual impairment on navigation ability. These studies also suggest that information provided by other senses or assistive technology can improve navigation ability with low vision.