Using AI for Accessibility: The OKO App and its Journey

In a recent episode of the AT Banter podcast - Ryan, Steve, Rob and and Lis - sits down with Michiel, the Founder and CEO of AYES, to discuss their groundbreaking application called OKO.

Listen here to the podcast:

https://atbanter.com/2023/07/05/at-banter-episode-342-michiel-janssen-oko/

OKO, a mobile app developed by AYES, was inspired by the challenges faced by their blind friend when navigating cities independently. The team noticed that these difficulties were not just local to Belgium but were a universal problem. Their solution? An app designed to make cities more accessible for blind and visually impaired individuals.

Making Cities More Accessible

Michiel highlighted the lack of Accessible Pedestrian Signals (APS), also known as audible signals, in many cities. AYES' mission is to offer an alternative in the form of a mobile application, OKO. Using the smartphone’s back camera and artificial intelligence, OKO interprets pedestrian signals, offering similar functionality to an APS. The app works when users hold their phones at chest level, aiming their back camera toward the pedestrian signal. Once the traffic light is within sight, the app immediately informs the user about the signal's status. In the United States, there are typically three signal statuses – walk, don’t walk, and countdown – and for each of these, OKO provides a different type of feedback.

The Different Types of Feedback

There are three types of feedback that OKO provides - an audible cue, a haptic or vibration cue, and visual feedback. The feedback is intuitive and mirrors the physical audible signal. For instance, when OKO detects a 'walk' signal, it provides a fast beep and a fast vibration. On the other hand, for a 'don't walk' signal, it offers a slow beep and a slow vibration.

Harnessing the Power of AI

The artificial intelligence utilised in OKO operates entirely on the smartphone, removing any dependency on Wi-Fi or a cellular connection. This means users can even go offline, needing only a charged battery to use the app. OKO also enhances users' orientation at intersections, helping prevent veering off into traffic. By holding the phone at chest level and gradually rotating their upper bodies towards the parallel street, users can orient themselves to the pedestrian signal at the other side of the road. However, OKO provides more information about the physical world, it's meant to be an additional tool, not a replacement for other Orientation and Mobility (O&M) skills, a white cane, or a guide dog.

Working with the Community

The AT Banter team noted two significant positives about the OKO application - its development involved working directly with the community, and it doesn't attempt to replace other O&M skills and devices. AYES originally considered developing a wearable device but decided against it due to the high computational resources, battery efficiency, and high-quality camera of current smartphones, particularly iPhones. This belief aligns with the AYES' mission to deliver natural, accessible solutions through devices that people already own.

Availability of the OKO App

Currently, the OKO app is only available in the US, Japan, Spain and Belgium. Interestingly, in Belgium, iPhones are considered tools that are reimbursed by the government, recognising their value for accessibility. The journey of AYES and its OKO app emphasizes the powerful impact of AI on accessibility. By providing a free, intuitive, and efficient solution to the real-world problem of urban navigation for blind and visually impaired individuals, the company sets an example of how technology can be harnessed for the greater good.