UC Santa Cruz

Navigating indoors without a map can be challenging, especially for visually impaired individuals in unfamiliar settings. Many existing indoor navigation methods rely on building maps, pre-deployed infrastructure (like BLE beacons or RFID), or visual-based systems that require a clear line of sight to a camera. These requirements can make technology less accessible to visually impaired individuals for independent navigation.To address these challenges, this thesis introduces the SafeReturn app, a new smartphone-based technology implemented as an iOS app designed to assist visually impaired individuals in returning to their starting point after navigating through indoor spaces. SafeReturn is particularly useful in real-life scenarios, such as hospitals, where a blind individual may initially navigate from a waiting room to a doctor's office with the assistance of a receptionist but subsequently needs to return independently. This technology eliminates the need for pre-deployed infrastructure or a clear camera view, enabling visually impaired individuals to navigate independently and confidently. The system features a new path-matching algorithm enhanced by a hybrid matching approach that integrates magnetic field and inertial data (representing steps/turns information) to backtrack paths. When active, it records sensor data and provides guidance for users seeking to retrace their route. Additionally, it includes an off-route detection mechanism that alerts users when they deviate from the correct path and provides notifications for path recovery to guide them back on track. Initial testing was conducted using the WeAllWalk dataset containing inertial data from blind walkers. Subsequently, the system was equipped with a watch-based user interface and speech-based notifications specifically designed to simplify interaction for blind users. A user study involving seven visually impaired participants at the University of Santa Cruz’s BE building demonstrated the effectiveness of the proposed localization solutions. The results from these tests illustrate the system's efficacy in assisting visually impaired individuals with indoor navigation and path recovery in real-world settings.