Smart Glasses for Visually Impaired People
Hadija Nyevu Said (team leader), George Nzomo, Husna Said, Nora Dama and Jesca Mteri, students at Mariakani Vocational Training Centre in Kilifi County, Kenya, developed Smart Glasses for Visually Impaired People in the Dariu Tech Club™ under the supervision of Moses Oduor, Dariu Master Trainer and centre principal. The spectacle-mounted device describes the wearer’s surroundings as spoken guidance. It took second place at all three Dariu Tech Club™ showcases held in 2026.
In Kenya, around 750,000 people are visually impaired, 250,000 of them live with blindness. Especially in rural coastal counties, reading a sign, crossing an unfamiliar street or recognising who has joined a conversation often depends on others. The white cane misses distant and overhead hazards, and other devices are usually bulky, costly or single-purpose. The team set out to build an “extra set of eyes”.
HOW IT WORKS
Ultrasonic sensors and a small camera feed spoken guidance through an earpiece:
- Obstacle detection: Ultrasonic sensors on the frame measure the distance to poles, steps, kerbs and low-hanging branches, at head height as well as at ground level. Warnings reach the wearer through a wired earpiece that leaves ambient sound audible.
- Object recognition and text reading: A forward-facing camera identifies what the wearer is approaching, using a model trained on labelled images of common obstacles. An optical character recognition (OCR) chip turns signs, labels and mail into speech.
- Face detection and name recall: The camera detects faces and announces by name those the wearer has registered, which helps in conversations and social settings.
- GPS navigation and voice assistant: A GPS module gives spoken turn-by-turn directions, and a voice assistant answers questions hands-free.
- Connectivity and power: Bluetooth and Wi-Fi link the glasses to a smartphone for navigation data, updates and configuration. A rechargeable battery powers the device.
IMPACT
Blind people gain help with navigation, reading and recognising faces, and people with partial vision loss receive audio guidance in addition to what they can still see. The current prototype demonstrates ultrasonic obstacle detection with spoken feedback; camera, object-recognition and OCR functions are the next stage, followed by trials with blind and low-vision users.



