Advancement: Safe Return: An Integrated Indoor Backtracking System For Visually Impaired People

Chia Hsuan Tsai
Computer Engineering PhD Student
Location
Virtual Event
Advisor
Roberto Manduchi

Join us on Zoom: https://ucsc.zoom.us/j/96768123072?pwd=bmJoaThYR2FKWHFmOFVrUnNGQlVWQT09 / Passcode: 349984

Description: Navigating in indoor environment without a map could be challenging and this is especially true when a visually impaired person is attempting to return to the starting point after walking along a certain path in an unfamiliar setting. The solution presented here is Safe Return, implemented as an IOS App, that provides guidance to assist visually impaired individuals to return to their starting point in an indoor environment. When the system is active, it records the sensor data which includes the magnetic field and inertial data provided by the smartphone. I also propose a new algorithm to integrate the magnetic field and steps/turns information to backtrack the path. Safe Return will generate directional information without relying on the map of the building or any pre-deployed infrastructure (e.g., BLE beacons or Wi-Fi AP). Furthermore, it provides an off-route detection mechanism and estimates the user’s real-time facing direction. When users go off-route, it will guide them to the correct walking path.

A preliminary version of the system was tested in the University of Santa Cruz’s E2 building and the result shows that using the magnetic field and steps/turns information could provide reliable guidance for backtracking the path. In future work, the system will be evaluated via user studies with visually impaired participants and further improvements on parameter estimation in the path matching algorithm will be conducted to enhance positioning accuracy.