From Pandemic Project to Award-Winning Innovation

The story of Oko began during the height of the COVID-19 pandemic in Antwerp, Belgium. Vincent Janssen, along with his brother Michiel and longtime friend Willem Van de Mierop, found themselves with newfound free time and a desire to apply their expertise in artificial intelligence to a meaningful problem. A close friend with low vision experienced increasing isolation during lockdowns, unable to rely on volunteer guides for safe navigation.

“We started to notice his increasing isolation,” explains Janssen. “The restrictions meant he couldn’t go for walks with a guide. That sparked the question: could we leverage AI, similar to the technology driving autonomous vehicles, to create a system that would help people navigate public spaces independently?”

Initially, the team experimented with “hacky hardware”—a combination of microcomputers, 3D-printed components, and a borrowed speaker—to test the feasibility of recognizing pedestrian signals. These early prototypes involved extensive field testing, meticulously analyzing the AI’s ability to identify signal states and perform reliably in various weather conditions. However, they quickly realized that a software-based solution offered greater potential for scalability and usability.

Despite having limited experience with iOS development, the trio embarked on a crash course in Swift, dedicating themselves to learning the necessary skills to bring their vision to life. “I had maybe opened Xcode three times before,” Janssen admits, “but we were determined to overcome that hurdle.”

The resulting app, Oko—meaning “eye” in several languages—utilizes the iPhone’s camera, Maps SDK, and a sophisticated AI algorithm to provide users with real-time feedback through haptic, audio, and visual cues. The app’s core functionality centers around identifying pedestrian signals and communicating their status—walk, don’t walk, and countdown timers—in an accessible format.

Early user testing revealed a surprising preference for portrait mode, prompting a significant redesign of the user interface. The team also refined the audio feedback to more closely mimic natural sounds and incorporated additional visual cues based on user input. This iterative process underscored the importance of co-creation with the accessibility community.

Launched in December 2021, Oko quickly gained recognition within the accessibility community. The team has since expanded its reach, conducting field research in the United States to adapt the app to different traffic patterns and pedestrian infrastructure. Today, the Oko team has grown to six, including dedicated developers focused on refining the app’s performance and adding new features.

What sets Oko apart is its unwavering commitment to simplicity. While the team explored potential expansions—such as integrating public transportation information—they ultimately decided to prioritize a focused, streamlined experience. “We realized that adding complexity would detract from the app’s core value,” Janssen explains. “Our goal is to provide a reliable and intuitive tool that empowers users to navigate their surroundings with confidence.”

The success of Oko demonstrates the transformative potential of AI when applied with empathy and a deep understanding of user needs. It’s a powerful reminder that technology can be a force for inclusion, breaking down barriers and creating a more accessible world for everyone. But what other everyday challenges could be addressed with similar AI-driven solutions? And how can developers ensure that accessibility remains at the forefront of innovation?

Frequently Asked Questions About Oko