News release

From: Springer Nature

AI smartphone app helps people with blindness navigate outdoors

A smartphone application named Mobilio that may help people with blindness or visual impairment to navigate outdoors more quickly and independently is reported in Nature Biomedical Engineering. Mobilio combines machine learning, sensor fusion, and personalised audio guidance to support independent outdoor navigation.

Blindness and visual impairment can restrict independent mobility and lead to reduced physical health, increased risk of collisions or falls, and lower quality of life. Although electronic travel aids are promising assistants for people with a visual impairment, they are often expensive or inaccessible, and may lack functions such as obstacle detection or guidance along walkable paths.

Patrick Slade and Raymond Liu present Mobilio, a smartphone app that works with a smartphone worn on the user's chest. The system uses the phone's camera, movement sensors, GPS, and light detection and ranging (LiDAR), together with open-ear headphones and a standard white cane. The system gives spatial audio cues in the form of beeps to guide users along routes and around obstacles, with personalised feedback. In tests, 9 participants with blindness and 5 participants with a visual impairment used Mobilio and a white cane to navigate an outdoor community path. The app was found to reduce navigation time by an average of 13% and required about 72% fewer verbal corrections, compared with navigation using Google Maps and a white cane. In an obstacle course, Mobilio reduced environmental contacts by an average of 41%, but increased course completion time by about 17%.

The authors note that further work is needed across more navigation tasks, locations, weather conditions and extended community use, as well as additional development of the user interface and wearable designs.