Smart Wheelchair IoT System

A comprehensive assistive technology solution designed to enhance the safety, independence, and health monitoring of elderly and specially-abled individuals using IoT, edge computing, and machine learning.

---

Overview

This Smart Wheelchair system integrates real-time fall detection, vital sign monitoring, and location tracking using multiple sensors, processed locally via the ESP32-WROOM-32 microcontroller and visualized through a Flutter-based mobile/web app. It features both immediate edge-side alerts and cloud-side intelligence for long-term predictive insights.

---

System Architecture (IoT Level 5)

• Perception Layer:

  • MPU6050 – Fall detection
  • MAX30100 – Heart rate and SpO2 monitoring
  • NEO-6M GPS – Real-time location tracking
  • Ultrasonic Sensor - Object Detection

• Network Layer:

  • ESP32 communication via MQTT/HTTP over Wi-Fi

• Data Processing Layer (Edge):

  • Preprocessing, noise filtering, and local alert logic

• Cloud Layer:

  • Firebase for real-time data storage, user auth, alerts, and ML-based analytics

• Application Layer:

  • Cross-platform Flutter dashboard for data visualization
  • Mobile app developed in Android Studio for real-time alerts and location tracking

---

Hardware & Software Stack

Sensors & Actuators

• MPU6050 – Motion/orientation sensor
• MAX30100 – Biomedical sensor for vitals
• NEO-6M GPS – Geo-tracking
• Ultrasonic Sensor - Object Detection
• Buzzer – Audible alerts
• ESP32-WROOM-32 – Edge device and comms hub

Software Tools

• Arduino IDE – Firmware development
• Flutter – Web application dashboard
• Android Studio - Mobile app development
• Firebase – Realtime DB, Auth, Functions, Hosting
• ML Models – Random Forest, XGBoost for predictive insights

---

Intelligence at the Edge & Cloud

At the Edge (ESP32):

• Multi-stage fall verification algorithm
• Local vitals threshold checks
• Offline alerting & data buffering
• Buzzer-triggered emergency response

At the Cloud:

• Firebase ML + Functions for:

  • Predictive health insights
  • Alert automation
  • Secure caregiver access
  • Visualization of ROC, trends, and model accuracy

---

Dashboards & UI

• Real-time heart rate and SpO2 monitoring
• GPS-based wheelchair tracking
• Historical trends and anomaly highlighting
• Mobile app notifications for critical events
• Role-based web access for caregivers
• Accessible UI (high contrast, speech-friendly)

---

Outcomes

• 95%+ accuracy in fall detection
• Continuous vitals tracking with secure cloud sync
• Real-time alerting within seconds of critical events
• ML-driven health risk predictions (heart-risk)
• Consistent performance in various testing environments

---

Getting Started

1. Clone the Repo

   git clone https://github.com/Prxyankaz/SMART-WHEEL-CHAIR.git

2. Hardware Setup

   • Connect the sensors and buzzer with ESP32
   • Mount the setup on a wheelchair

3. Flutter App

   • Open in VScode
   • Install all dependencies
   • Configure Firebase credentials
   • Run and deploy

4. Mobile App

   • Open in Android Studio
   • Connect Firebase (Ensure google-services.json is present in the app/ directory and matches your Firebase project)
   • Sync Gradle & Run

5. Web Interface

   • Deploy to Firebase Hosting or local server

6. Mobile app Interface

   • Connect a Device or Emulator - Launch a virtual device via AVD Manager, or connect your Android phone with USB Debugging enabled.

---

Contributors

• Harini Priyanka W (CB.EN.U4CSE22018)
• Mogitha S M (CB.EN.U4CSE22027)
• Prahalyaa A (CB.EN.U4CSE22432)

---

License

This project is licensed under the MIT License – feel free to build upon it for academic or nonprofit purposes.

---