iKshana
A complete navigational solution for the visually impaired
Institute Technical Summer Project, IIT Bombay
Role: Team Lead
Duration: May - Aug 2023
Achievement: 2nd Place among 100+ teams
The Beginning
It started with a simple question during my first summer at IIT Bombay: How do visually impaired people navigate unfamiliar places? White canes detect obstacles but can’t tell you which way to turn. Guide dogs are expensive and unavailable to most. GPS apps speak directions, but audio can be drowned out by traffic noise or draw unwanted attention.
We had three months, zero hardware experience, and an idea that seemed almost naive: what if your shoes could guide you?
The Design
iKshana—Sanskrit for “vision”—became a foot-worn harness with two complementary systems:
Short-range navigation using ultrasonic sensors detecting obstacles within 30cm. When you’re about to walk into something, the corresponding foot vibrates. Left foot for obstacles on the left. Right foot for the right. Both feet for something directly ahead.
Long-range navigation using GPS and an Android app we built from scratch. We processed OpenStreetMap data to identify every intersection on the IIT Bombay campus. At each turn, the appropriate foot pulses three times—left for left turns, right for right. Both feet together means keep walking straight.
The hardware consisted of an ESP-32 microcontroller on each foot, communicating with the Android app via WiFi. Each module had two ultrasonic sensors (HC-SR04) pointing forward and to the side, plus coin vibration motors for haptic feedback. Everything was powered by compact Li-Po batteries and housed in custom 3D-printed enclosures.
The Software
The Android app was the brain of the operation. It handled three critical tasks:
- Location tracking using the phone’s GPS and fused location provider
- Route calculation using processed OpenStreetMap data with the Osmium library
- Communication with the ESP-32 modules to trigger appropriate vibration patterns
We preprocessed the entire IIT Bombay campus map, extracting intersection coordinates and path data into JSON format that the app could quickly query. When the user approached an intersection, the app calculated the required turn direction and sent commands to the appropriate foot module.
How It Works for the User
From the user’s perspective, iKshana feels like a silent guide always walking with you. You open the app, select your destination on the campus map, and start walking as usual. At each turn, the corresponding foot vibrates three times—left foot for a left turn, right foot for a right turn, both feet together when you should keep going straight. When an unexpected obstacle appears within 30cm, the system warns you with a sharper vibration on the side of the obstacle.
Building It
We learned everything on the fly. Soldering irons burned fingers. ESP32 WiFi connections dropped mysteriously at 3 AM. Our first 3D-printed enclosures cracked under the stress of walking. The Android app crashed more often than it worked.
But slowly, it came together. The vibration patterns became intuitive. The GPS accuracy improved. The enclosures got stronger. By August, we had two working modules that could genuinely guide someone across campus.
The Result
We presented iKshana at the ITSP finale in front of faculty, industry judges, and hundreds of fellow students. Walking them through a blindfolded demo—watching their faces as they realized the vibrations actually worked—was unforgettable.
Second place among 100+ teams. More importantly, we’d built something real. Something that worked. Something that could actually help people.
What I Learned
This was my first hardware project. I learned to solder, design PCBs (badly, at first), write Android apps, process map data, and debug systems where the bug could be in the code, the wiring, the sensor, or the battery simultaneously.
But the real lessons were about teamwork. About pushing through when nothing works. About the feeling when it finally does. The friendships forged in Tinkerer’s Lab at 4 AM, debugging code that should have been working, became some of the strongest I have.
Presentation
Repository: github.com/abhineet-agarwal/ikshana