Inspiration
We wanted to make assistive navigation more affordable and accessible using everyday technology. AHEAD was born from the idea that understanding your surroundings shouldn't require expensive specialized equipment.
What it does
AHEAD combines AI-powered object detection, iPhone LiDAR, spatial memory, voice descriptions, and wearable haptic feedback to help people with visual impairments understand obstacles and possible directions.
How we built it
We connected an iPhone to a Python FastAPI backend running OpenCV, YOLO, and spatial navigation algorithms. ARKit provides depth and camera tracking, while an ESP32 and SG90 servo translate directional decisions into tactile feedback.
Challenges we ran into
Handling reflective surfaces, small obstacles, unreliable depth readings, changing camera angles, and unstable directional decisions while keeping the entire system responsive.
Accomplishments that we're proud of
Built an end-to-end working hardware prototype with real-time LiDAR navigation, spatial memory, and voice interaction. Passed 161 software tests and evaluated 3,600 simulated depth frames.
What we learned
Reliable assistive AI needs more than object recognition. Spatial reasoning, sensor fusion, latency, accessible feedback, and knowing when not to give a direction are equally important.
What's next for AHEAD
Expand real-world validation, improve obstacle detection and wearable ergonomics, and work with visually impaired users to refine the experience.
