Empowering independence through intelligent assistive technology
Presented by
Ojasvi Mishra, Akhil Mor, Nitin Mishra, David Lakra
Lovely Professional University, Punjab-144411, India
UID: 12324518
ojasvimishra9792@gmail.com
+91 9044256927
UID: 12307177
busysan125@gmail.com
+91 83076 03629
UID: 12306101
nitinmishra979318@email.com
+91 95802 40030
UID: 12323849
lakradavid396@gmail.com
+91 7205187469
Lovely Professional University, Punjab-144411, India
Over 285 million people worldwide are visually impaired, with 39 million completely blind. Navigation remains one of their biggest daily challenges.
Only detects obstacles at ground level within 1-2 meters. Cannot identify overhead hazards, moving objects, or provide directional guidance.
High risk of collisions with obstacles outside cane range. Difficulty navigating unfamiliar environments independently.
Limited detection range of traditional white cane
Blind spots: Overhead, beyond 2m, fast-moving objects
JAWS, NVDA, VoiceOver provide digital accessibility but no physical navigation support.
OrCam, Envision AI offer camera-based assistance with voice feedback.
WeWALK, SmartCane add ultrasonic sensors to traditional canes.
Key Gap: No existing solution combines multi-sensor fusion, AI-powered object detection, GPS navigation, and affordable pricing in a portable, attachable form factor.
A compact, sensor-rich device that attaches to any standard white cane or can be worn independently, providing comprehensive environmental awareness through AI-powered analysis.
Compact design with multi-sensor array
System Architecture
Inside Dissection
Probable Look of the Product
Combines LiDAR, ultrasonic, camera, GPS, and IMU for comprehensive environmental perception. No single point of failure.
GPS-enabled turn-by-turn guidance with dynamic route adjustment based on obstacles and user preferences.
Dual feedback system ensures alerts are never missed. Voice guidance for navigation, vibration patterns for obstacle proximity.
Under 200g total weight. Universal clamp attaches to any cane. Can also be worn as a chest-mounted device.
Machine learning classifies obstacles and predicts movement patterns. Recognizes traffic signals, crosswalks, and pedestrian zones.
3000mAh battery provides 12+ hours of active use. USB-C fast charging. Power-saving modes extend operation.
Each detected object is annotated with:
Builds an annotated virtual map of the environment.
NLP-based speech synthesis converts detected objects and attributes into audible information:
Touch-sensitive interface maps screen areas to physical object positions for tactile exploration.
3D representation continuously refreshed using Euler angle rotations (yaw, pitch, roll) and IMU translation vectors. Virtual environment stays aligned with the user's spatial reference.
System updates automatically when users reposition the device or capture new images. Handles dynamic obstacles like moving people and vehicles.
Advanced sensor fusion via Extended Kalman Filter enhances accuracy and spatial consistency by optimally combining noisy sensor readings.
Simultaneous Localization and Mapping builds a persistent map of the environment while tracking the user's position within it in real time.
Users can store the location of important objects (keys, wallet, medicine) and request step-by-step guidance to retrieve them later.
AI path optimization algorithms provide directional guidance step-by-step using stored spatial information and real-time obstacle data.
Seamless transition between outdoor GPS navigation and indoor IPS using Bluetooth beacons and RFID for complex environments like malls and transit hubs.
IoT-enabled pedestrian infrastructure: adaptive traffic lights, smart crosswalks, and geofenced pedestrian zones provide audio and haptic crossing alerts.
Crowdsourced object annotations, AI model updates, and navigation preference customization via cloud. 5G and edge computing enable low-latency data exchange.
Swipe gestures to rotate the 3D scene. Voice commands for hands-free operation. NLP models (GPT-based) enable conversational navigation queries.
Each sensor module is independently replaceable and upgradable. Users can customize based on their specific needs and budget.
Unlike proprietary solutions, our device attaches to any standard white cane or wheelchair without modifications.
Patent-pending dual-channel feedback combines spatial audio with haptic patterns for intuitive obstacle awareness.
All AI inference runs locally on-device. No internet required for core functionality. Privacy-preserving by design.
Integration with smart traffic signals, accessible pedestrian signals, and city-wide navigation networks.
Bus stop detection, train platform navigation, and real-time transit updates via app integration.
Indoor navigation, store finder, and obstacle avoidance in crowded retail environments.
Gate navigation, boarding announcements, and assistance request features for stress-free travel.
Integration Potential: API available for third-party developers. Compatible with Google Maps, Apple Maps, and OpenStreetMap.
Secures early-stage innovation for 12 months — covers AI-driven 3D mapping, sensor fusion, real-time navigation, and smart feedback integration.
Full legal protection with detailed technical claims — 20 years of exclusivity.
Patent Cooperation Treaty filing extends priority up to 30 months for multi-jurisdiction commercialization.
Filed under Indian Patent Act, 1970 with fast-track options under Startup India & MSME schemes.
Develop companion smart glasses with bone-conduction audio and heads-up display for enhanced awareness.
Natural language voice assistant for conversational navigation queries and contextual recommendations.
V2X communication for safe autonomous vehicle boarding and coordination with ride-sharing services.
Our sensor-integrated attachable navigation device represents a significant leap forward in assistive technology for the visually impaired community.
Bridges the gap between traditional mobility aids and expensive high-tech solutions
Enables true independence through intelligent, reliable navigation assistance
Scalable, affordable solution with potential for global accessibility impact
Assistive technology is not just about devices; it's about dignity, independence, and equal opportunity.
A wireless device for blind people with voice-activated control, object scanning, memory storage, journey planning via GPS, and medical monitoring capabilities. Combines communication, scanning, and identification features for multifunctional independence.
A mobile device technique for blind users using Multi-View Stereo and Structure from Motion to create a 3D representation of the environment. Users explore via touchscreen with objects linked to dimensional attributes. 3D image is sensitive to device orientation.
A guide system using multiple sensors to identify a clear path for ambulatory vision-impaired persons. Includes wheels, a platform with processor, a rigid harness with haptic feedback grip, and sensors to sense the environment. Processor identifies objects and sends avoidance messages.
Collaboration with Google Lookout, Microsoft Seeing AI, and Aira for text, object, and currency recognition. GPT-4 NLP models for conversational accessibility.
Shopping malls, airports, and transit centres using BLE beacons and RFID positioning. AI-enhanced real-time transit tracking for buses, trains, and ride-sharing.
Integration with haptic wearables: smart cane, vibrating footwear, AI-controlled gloves. AI gesture recognition for command inputs.
AI pedestrian recognition in self-driving ecosystems. V2X (Vehicle-to-Everything) communication for safer road crossing and autonomous vehicle boarding.
Federated learning for personalized AI training without compromising user privacy. Predictive analytics for navigation efficiency based on real-time traffic and weather.
Framework expandable to AR smart glasses, AI robotic navigation assistants, and wearable haptic sensors. Deployable on smartphones, embedded systems, and cloud-connected AI hubs.
Real-time 3D models via computer vision, inertial sensing, and AI depth estimation. Optical flow + stereo matching provides robust object localization.
Faster R-CNN, YOLO, and Mask R-CNN identify and label objects with contextual features: size, distance, and material properties for enhanced scene perception.
Sensor fusion with accelerometer, gyroscope, LiDAR, and ultrasonic improves positioning accuracy. SLAM enables dynamic path planning for indoor and outdoor settings.
Voice AI assistants, haptic feedback, and spatialized audio clues provide seamless interaction. Binaural sound localization helps users understand object positions.
Machine learning models enhanced with real-world interactions. Cloud repository enables automatic updates, crowdsourced annotations, and customized navigation preferences.
IoT-enabled adaptive traffic lights and smart crosswalks provide audio and haptic alerts. Retail and public space navigation via IPS with Bluetooth beacons and RFID.
for your attention
Ojasvi Mishra, Akhil Mor, Nitin Mishra, David Lakra
Lovely Professional University