Blind people using a white cane have limited walking speed because they have to wait for the tip of the cane to hit an obstacle before they can get around it. Unlike the NextGuide cane, it pre-emptively guides users around obstacles they haven’t yet reached.
Students of ETH Zurich student Alexander Bayer were inspired to invent the device after seeing how a blind classmate struggled with a traditional white cane. He is now commercializing the technology with three other students through ETH spin-off company NextGuide.
The cane itself consists of a collapsible carbon fiber pole connected to a handle unit that contains a 160-degree forward-facing camera, a haptic actuator, a physical tactile pointer that swings left or right, a microprocessor and other electronics .
As the user walks with a cane, the camera scans the environment in front of them. Machine learning algorithms on the processor analyze the images about 100 times per second, checking not only for obstacles, but also for doors, stairs and crosswalks.
When an oncoming obstacle is detected, the handle vibrates to get the user’s attention. Additionally, its pointer wiggles in the direction they’re supposed to walk to avoid obstacles—users feel the pointer move as their thumb keeps resting on it. Once the road ahead is clear again, the pointer guides them back on track.
Additionally, when the camera spots a door, staircase or crosswalk in front of the user, the cane will alert them of its presence through different vibration cues in the handle. Users can also select an audible alert if desired.
There’s no word yet on when the NextGuide cane will go into production, but potential buyers can sign up for updates by Company website. The cane is demonstrated in the video below.
NextGuide — Smart Walking Cane for the Blind and Visually Impaired