Harvard App Guides Blind Users Faster, Safer

A Harvard study found its Mobilio smartphone navigation app helped visually impaired users reach destinations faster and avoid indoor collisions.

Updated on Nov. 1, 2026 in Artificial Intelligence

Harvard App Guides Blind Users Faster, Safer

Live Poll

Should public funding support smartphone apps that help blind people navigate safely?

Harvard researchers published a study in Nature Biomedical Engineering on Mobilio, an AI smartphone app that gives real-time navigation and obstacle guidance to blind and visually impaired users. In a 14-person test, Mobilio users reached destinations 13% faster than cane users relying on Google Maps.

Why it matters

Google Maps is not detailed enough for safe use by visually impaired people, and Mobilio was built around needs identified through consultation with more than 100 people with limited or no sight. Because it runs on a standard smartphone without expensive specialized equipment, it could widen access to safer navigation tools across the United States.

The results come from a 14-subject controlled comparison, with Mobilio users showing 13% faster arrival times and 41% fewer indoor collisions than cane users relying on Google Maps. Performance in everyday, unstructured environments has not yet been measured.

The players

Harvard University

Harvard University is a private Ivy League research university in Cambridge, Massachusetts, and the developer institution behind the Mobilio navigation app.

John A. Paulson School of Engineering and Applied Sciences

The John A. Paulson School of Engineering and Applied Sciences is Harvard's engineering school, which is leading the planned real-world testing of Mobilio.

The details

Mobilio uses a smartphone's camera, GPS, inertia detection and LiDAR when available, paired with a computer-vision model that analyzes live video to build a navigation plan from the user's location to a chosen destination. The app delivers customized audio cues for continuous guidance, turn-by-turn directions and obstacle avoidance.

Study published in Nature Biomedical Engineering, with coverage dated November 1, 2026.
Harvard's SEAS team plans further real-world testing to assess how Mobilio fits users' daily routines and preferences.

Big Picture

Mobilio is positioned as a specialized alternative to Google Maps, which researchers found lacks the granularity needed for safe pedestrian navigation by blind and visually impaired users. Its reliance on pedestrian-based semantic segmentation datasets, rather than driver-focused data, marks a deliberate departure from consumer mapping tools built primarily for vehicles.

Because Mobilio runs on a standard smartphone without expensive specialized hardware, blind and visually impaired users may be able to adopt it without buying costly equipment. Its current status is a research study, so no public release date or purchase details have been announced.

The takeaway

For visually impaired users, the study suggests that purpose-built, pedestrian-focused navigation can outperform general mapping tools in both speed and safety. Readers tracking assistive technology should watch for the planned real-world trials that will test whether these gains hold outside the lab.