According to a press release, students at the Swiss Federal Institute of Technology Zurich (ETH) have made a smart cane using computer vision and machine learning. They integrate a camera, software, and a pointer into their so called NextGuide. The blind individual can feel tactile signals through their thumb to receive instructions on where to turn to avoid obstacles.
Arvid Gollwitzer, one of the inventors, explains in a video: “Conventional white canes only enable this by bumping into an obstacle and inching forward slowly. With our cane, blind individuals can move around obstacles fluidly in a very natural way.”
According to the provider of the idea, Alexander Bayer, another advantage is found in pedestrian crossings, which cannot normally be felt with a cane. These are registered by the built-in camera. This applies to doors as well. If the integrated technology detects a door, the cane handle vibrates twice and the pointer indicates its direction.
ETH student Alexander Bayer was inspired by a former blind classmate. He was able to “use his ETH knowledge for the project” at the Student Project House and found collaborators there. The ETH students’ NextGuide is “already far into its development” according to the German Society for Deaf-Blindness. “They hope to bring it to market in one to two years.”
