A little past noon on a recent autumn day, Hiroyuki Yoshizumi demonstrated his reliance on a special app to help him safely navigate his surroundings.

In early October, Yoshizumi, 52, a resident of Fukuoka city’s Sawara Ward, was returning home from his job as a teacher at a Braille school.

Heading toward Tenjin Station, about a 10-minute walk, Yoshizumi held his smartphone to his ear and started Eye Navi.

The app quickly announced, via audio guidance, not only the locations of a convenience store, a post office and other nearby establishments but also their distances in succession.

“A Lawson convenience store has just opened right here,” Yoshizumi said, smiling with delight at his discovery.

Yoshizumi wielded a white cane to move along tactile paving blocks. He explained that he often is startled when he is about to step onto the road or crashes into a bicycle, breaking the stick.

“I always feel my life is at risk,” Yoshizumi said. “Taking an aimless stroll is like a pipe dream for us, visually impaired people.”

The Eye Navi audio guide app is increasingly emerging as a reliable guide for visually impaired individuals such as Yoshizumi.

The app takes full advantage of artificial intelligence to detect people, cars, traffic lights, utility poles and other signs along streets.

Eye Navi reportedly guides users effectively through voice instructions and has seen its usage grow particularly in recent years.

Recognizing “automobiles,” “poles,” “passers-by,” “car stops” and “red lights” in succession, Eye Navi on the smartphone in his chest pouch auditorily promptly shared what appeared just in front of Yoshizumi.

The AI system similarly tells the user “which direction is right,” so Yoshizumi said he is “relieved” thanks to the kind announcement on his walks.

Braille blocks and audible traffic signals were not as ubiquitous as they are today, when Yoshizumi became completely blind. They have since been incorporated into society in phases.

Despite this, people living in the neighborhood at times complain about the “noise” from such traffic lights and request they should be muted at night and early in the morning.

The increasing presence of gas-electric hybrid vehicles renders it much more difficult for individuals with poor vision to detect automobiles by the sound of their engines.

Those with visual impairments are therefore still plagued by persistent anxieties during their outings.

Toward the end of 2023, Yoshizumi began relying on Eye Navi at the recommendation of an acquaintance. Yoshizumi can now rush over whenever invited to a drinking party. He can similarly reach his favorite bar smoothly and safely now on his own.

“I was happy that I could get home safely by myself even if I was drunk,” he recalled.

VISION FOR HIGH-MOBILITY SOCIETY

Eye Navi was developed by Computer Science Institute Co., headquartered in Kita-Kyushu.

The corporation’s president, Hidemi Hayashi, 74, is well known as a former vice president at the major mapping company Zenrin Co.

At Zenrin, Hayashi spearheaded the digitization of maps and its business involving car navigation systems.

Eye Navi came into existence in line with the aim of “realizing a society in which everyone can move about freely and enjoyably anywhere.”

But the initial idea was pitching a “robotic guide dog” to accompany citizens suffering from vision problems.

After its founding in 2015, Computer Science Institute conducted repeated verification tests based on the opinions and feedback from visually impaired individuals. The robot dog project was eventually abandoned due to the extremely high technical hurdles.

Seeing advances in AI technology dramatically expand the potential of the envisioned guide system for the visually impaired around that time, Computer Science Institute changed plans.

It then completed a high-precision mechanism that informs the user by voice of what is ahead, after the AI model was trained to recognize 20 types of high-priority targets along roads, inclusive of cars, bicycles and traffic lights.

Eye Navi hit the market in 2023 and has already been downloaded more than 33,000 times.

Cashing in on Zenrin’s sponsorship, the basic features of Eye Navi, such as route guidance, obstacle detection and automatic image recording during walks, are available free of charge.

Still, the Computer Science Institute is only halfway to realizing its ultimate policy and vision.

It is difficult for Eye Navi to detect a downward step in the direction of travel. Also among challenges to overcome in the future are location errors that occur in urban settings with a lot of high-rise buildings under GPS.

With these problems in mind, Hayashi expressed hopes for the future.

“The project was initially launched with the aim of helping people with visual impairments,” Hayashi said. “We will be continuously committed to stepping up the app’s detection accuracy further, adding new functions to enrich their lives at the same time.”