What it does
Smart Vision Glasses for the Blind：This AI-powered assistive glasses converts visual info into vibrations & voice prompts, solving navigation challenges. Its IR/camera system detects obstacles (0.5-5m), while GPS/UWB enables precise indoor/outdoor navigation.
Your inspiration
Inspiration Behind the Design ： The core inspiration came from observing the daily struggles of visually impaired individuals navigating unpredictable urban environments. While existing solutions like white canes and basic GPS devices provide partial assistance, they fail to deliver real-time, comprehensive environmental awareness. The breakthrough idea emerged when we recognized how modern sensor fusion technology could replicate the brain's natural multi-sensory integration - where vibration patterns could serve as tactile "pixels" and AI voice as an interpretive layer.
How it works
**Technical Operation Principles 技术运作原理** **1. Environmental Sensing System 环境感知系统** The device employs a hybrid sensor array: - Dual 8MP stereoscopic cameras (3280×2464@15fps) with HDR imaging - 940nm VCSEL infrared ToF sensor (Resolution: 640×480, Max range: 8m, Accuracy: ±1%) - 6-axis IMU (±16g accelerometer + ±2000dps gyroscope) for motion compensation - Ambient light sensor (1-100,000 lux detection) These components work synergistically: 1) ToF measures absolute distance while stereo cameras provide texture details
Design process
**Phase 1: Problem Identification (问题定义阶段)** Conducted 3-month ethnographic research with Beijing Association of the Blind, discovering key pain points: - 78% of participants reported collision incidents with low-height obstacles (<1m) - Existing electronic travel aids averaged 2.3x heavier than comfortable wear threshold - Voice interfaces caused 2.8-second average decision delay in noisy environments **Phase 2: Conceptualization (概念生成阶段)** Developed 12 concept variants through: 1) **Biomimicry Study**: Adapted bat echolocation principles into IR pulse patterns 2) **Hardware Prototyping**: - Iteration 1: Raspberry Pi + Intel RealSense (rejected due to 380g weight) - Iteration 2: Custom PCB with OAK-D-Lite spatial AI camera (reduced to 210g) 3) **Feedback Mechanism Testing**: - Compared 5 vibration encoding schemes, selecting "frequency+duration" combo **Phase 3: Form Development (形态开发阶段)** Created 3D-printed prototypes addressing: - **Weight Distribution**: Relocated battery from temple to rear headband
How it is different
**Innovative Differentiation 创新差异点** **1. True Multi-Sensory Integration 真正的多感官融合** Unlike conventional single-mode aids (vibration-only or voice-only), our system implements: - **Context-Aware Feedback Switching**: Automatically selects optimal output based on environment noise level (e.g. vibration dominates in 80+dB noise) - **Cross-Modal Validation**: Voice confirms vibration alerts only when user remains stationary >2s *Competitor Gap*: Existing products force manual mode switching, causing cognitive load **2. Dynamic Obstacle Modeling 动态障碍物建模** Patent-pending algorithm combines: - Short-term prediction (LSTM neural network forecasts moving objects' trajectories) - Risk quantification (calculates collision probability every 200ms) *Competitor Gap*: Rival devices treat all obstacles equally regardless of motion state
Future plans
**Future Development Plan 未来发展计划** **1. Enhanced AI Capabilities 增强AI能力** - Implement **multi-lingual real-time translation** (covering 10+ languages by 2025) - Develop **predictive navigation** using crowd-sourced accessibility data - Introduce **haptic feedback gloves** for tactile mapping of complex spaces **2. Medical-Grade Certification 医疗级认证** - Pursue **FDA Class II medical device** approval (target 2026) - Clinical trials with 300+ participants across 5 countries - Integrate with **electronic health records** for vision rehabilitation tracking **3. Ecosystem Expansion
Awards
江西省大学生科技创新竞赛工业设计赛项二等奖 全国大学生数字媒体科技作品及创意竞赛省赛三等 第六届香港当代设计大赛铜奖 全国青年大学生数字国防教育创新创意大赛省赛三等
Share this page on