“Create what’s essentially a small autonomous vehicle plan,” says Patrick Slade, the Harvard bioengineering professor behind a new navigation tool. He means a phone app, not a car.
- Mobilio is a research prototype from Harvard that uses a smartphone’s sensors to plan a safe walking path for people who are blind or have low vision.
- In a 14-person study, it beat Google Maps turn-by-turn directions on one outdoor route: about 13% faster, with about 72% fewer verbal corrections.
- You can’t download it. There’s no release date and no price.
How Mobilio finds a path
The idea is the same one a self-driving system uses to plan a route, scaled down to a pocket. For someone who is blind or has low vision, an outdoor trip means tracking curbs, bike racks, and crossings. A white cane catches what’s near the ground. Mobilio tries to add route guidance and obstacle avoidance on top of the cane.
Harvard’s engineering school says the app uses the phone’s camera, GPS, motion sensor, and LiDAR when the phone has it. A custom vision model picks out sidewalks, crosswalks, roads, and other walkable surfaces. The team had to build its own model because most public street datasets are captured from cars. The system combines that view with GPS and the user’s actual motion to plan a route.
It doesn’t talk you through each step. Instead it plays continuous beeps that come from the left or right to signal steering. Harvard says the team tuned the cues to each user by measuring how well they followed them. Nature Asia describes the phone as worn on the chest, with open-ear headphones and a white cane.
Raymond Liu, a Ph.D. student, led the research with Slade, and the work appeared in Nature Biomedical Engineering. Harvard says the design followed a survey of more than 100 blind or low-vision people. You can read Harvard’s official announcement.
What the study found, and what’s next
Fourteen volunteers from the Carroll Center for the Blind in Newton, Massachusetts, took part. Nature Asia says nine had blindness and five had visual impairment. Everyone used a cane. The comparison was Google Maps turn-by-turn plus a cane, against Mobilio plus a cane.
On the outdoor route, Mobilio users finished about 13% faster (Harvard SEAS and Nature Asia) and needed about 72% fewer verbal corrections (Nature Asia). In the indoor obstacle course they made about 41% fewer contacts (Harvard SEAS and Nature Asia), but took about 17% longer to finish (Nature Asia). Harvard also says Mobilio got users to their destination without major errors about as reliably as a human guide. News-Medical points out that the article doesn’t define “major errors.”
To put 13% in scale, take a hypothetical 20-minute commute. That’s about 2.6 minutes saved. It’s our own arithmetic on the study average, not a result for any real rider.
Mobilio is a research prototype. Harvard says the goal is a standalone app that runs entirely on the phone and could one day be downloaded from an app store. Harvard’s Office of Technology Development is pursuing commercialization, and the Harvard Grid Accelerator is helping turn it into a product. The team plans testing in more realistic settings around Boston.
Why It Matters to You
If you’re blind or have low vision, the promise is route guidance and obstacle avoidance on top of a white cane, using nothing but a smartphone. That isn’t available today. The early results suggest the idea is worth following, and the planned Boston testing should tell us more about how it holds up outside a small study.
Read the results with care
- Fourteen people from one organization is a small group. The authors say more testing is needed across tasks, locations, weather, and longer community use.
- The 13% is an average from 14 people on one outdoor route. It shows promise but doesn’t predict results for everyone.
- The only comparison was Mobilio versus Google Maps, both with a cane. Other navigation aids and guide dogs weren’t tested.
- Indoors, participants hit fewer obstacles but took about 17% longer to finish.
- The authors say the interface and wearable design need further work.
- News-Medical says the summaries don’t report statistical significance or participant demographics. The journal paper has the full methods.
Mobilio is a research tool. It isn’t a replacement for a cane, a guide dog, or orientation and mobility training.
Can I download Mobilio now?
No. It’s a research prototype. Harvard describes a future app that could be downloaded from an app store but gives no date.
Does it replace a white cane?
No. Participants used a cane in both conditions. Mobilio was tested as an add-on.
Do I need special equipment?
Harvard says it runs in real time on a standard smartphone and uses LiDAR when the phone has it. The study setup used a chest-worn phone and open-ear headphones, per Nature Asia.
Sources
- Harvard SEAS: A Smartphone Navigation App for People With Blindness and Low Vision (Aug 24, 2026)
- News-Medical: Harvard engineers develop smartphone app to aid navigation for people with blindness (Aug 24, 2026)
- Nature Asia: Press release on the Mobilio study (Aug 25, 2026)
We did not open the full journal paper. Study details come from the three pages above.
