About the Project — VIGIL: Smart System for Remote Diabetes Monitoring

🌟 Inspiration

The inspiration for VIGIL came from a real and serious problem: many diabetic patients experience sudden drops or spikes in glucose without noticing, especially when they are alone. This can lead to life-threatening diabetic comas that could be prevented if someone — or something — could detect the danger early.

I wanted to build a system that protects patients remotely, without finger pricks, and alerts the people who can save them.

🧠 What I Learned

During the project, I learned:

How non-invasive glucose monitoring works using technologies like infrared light, optical (PPG) sensing, and interstitial fluid (ISF) readings.

How to design a full AI prediction model that reads patterns and predicts danger.

How to turn a medical problem into a complete hardware + software solution.

How to create a professional pitch deck, select data, and simplify complex ideas.

How remote monitoring systems work in the real world.

⚙️ How I Built the Project

1. Hardware Concept

I studied how glucose levels in interstitial fluids (ISF) reflect blood glucose levels, and how sensors can read this using light waves — all without any needle.

A simplified physiological idea:

𝐺 𝐼 𝑆 𝐹 ( 𝑡 ) ≈ 𝐺 𝑏 𝑙 𝑜 𝑜 𝑑 ( 𝑡 − Δ 𝑡 ) G ISF ​

(t)≈G blood ​

(t−Δt)

Where Δ 𝑡 Δt is the delay between ISF and blood glucose changes.

2. Software / Application

The app was designed to:

Show real-time glucose levels

Display trends over time

Notify the user if their sugar rises or drops abnormally

Log historical readings in a clean interface

3. AI Prediction Model

The AI model analyzes glucose behavior and predicts hypoglycemia or hyperglycemia early.

General prediction logic:

𝑅 𝑖 𝑠 𝑘 ( 𝑡

)

𝑓 ( 𝑑 𝐺 𝑑 𝑡 , 𝐺 ( 𝑡 ) , past trends ) Risk(t)=f( dt dG ​

,G(t),past trends)

This helps the system warn the patient before the danger actually happens.

4. Emergency Alert System

If the app sends an alert and the user doesn’t respond:

The system repeats the alert.

If still no response → it sends the live location to:

Family

Doctor

Nearest ambulance

This ensures help arrives before the patient enters a coma.

5. Community & Doctor Support

The project also includes:

A Community Support Space for diabetic patients to share advice

Two 15-minute doctor consultations per month as part of the subscription plan

🧱 Challenges I Faced

The most important challenges:

Simplifying complex medical ideas for a presentation

Choosing a realistic non-invasive method that makes sense technically

Designing an AI model that is believable and functional

Organizing a professional pitch deck

Collecting accurate statistics from ADA & NIDDK to support the problem statement

Each challenge made the final result better.

💡 Final Thoughts

VIGIL is not just a device or an app. It is a Smart Remote Monitoring System designed to prevent diabetic comas before they happen.

It combines:

Continuous needle-free glucose monitoring

AI prediction

Smart alerts

Emergency automatic location sharing

Food recommendations

Community support

Doctor consultations

This project allowed me to combine medicine, artificial intelligence, hardware, software, and design into one unified system aimed at saving lives — not just tracking numbers.

Built With

• flattr