👁️‍🗨️ ElderEye: Family-Centered IoT Smart Care System

Real-time elderly monitoring platform combining live video streaming and anomaly detection
Providing peace of mind for families and caregivers — anytime, anywhere.

📌 Overview

ElderEye is a modular smart care system designed to continuously monitor the safety and well-being of elderly family members.
It seamlessly integrates real-time video streaming, deep learning-powered behavior anomaly detection, and instant mobile alerts to act as a reliable "eye and ear" for families and caregivers.

Built on a Hybrid Microservice Architecture, ElderEye ensures scalability, modularity, and robust observability.
Each service can be deployed independently, with comprehensive system monitoring enabled via Prometheus and Grafana.

---

⚙️ System Architecture

This architecture diagram illustrates the interaction among the mobile frontend, backend microservices, gRPC-based machine learning inference server, and IoT devices deployed in the elderly’s environment.

---

🚀 Key Features

• [1] Family-Centered Real-Time Video Streaming: Low-latency, multi-party video streaming using Mediasoup SFU to keep family members connected.
• [2] Behavior Anomaly Detection: Integration with an external deep learning model via gRPC for intelligent anomaly recognition.
• [3] Persistent Event Logging: Reliable CRUD operations and event storage powered by MySQL.
• [4] Instant Mobile Alerts: Timely notifications to caregivers through a React Native mobile app.
• [5] Modular Microservices: Independent deployment and scalability of backend services.
• [6] Full Observability: System health and metrics tracking with Prometheus and Grafana.
• [7] Data Privacy: Sensitive personal data securely encrypted at rest in the database using AES-256.
• [8] Authentication: User authentication and authorization managed via JWT (JSON Web Tokens).
• [9] Access Control: Principle of Least Privilege (POLP) enforced to restrict permissions to only what is necessary for each user and service.

---

🛠️ Technologies Used

• Backend: FastAPI, Mediasoup, gRPC
• Frontend: React Native
• Database: MySQL
• ML: PyTorch (pre-trained model integrated for inference; no training performed)
• Monitoring: Prometheus, Grafana

---

📸 Screenshots

App Home Screen	Live Streaming Screen	Alert Tab Screen	Notification Detected Screen
Home Dashboard	Real-Time Video Streaming	Alert Tab View	Detected Notification View

---

🗂️ Directory Structure

📁 ElderEye/
├── 📁 BE/                   # 🖥️ Backend services
│   ├── 📁 server1/          # 🧾 FastAPI – CRUD operations, event logging, alerts (MySQL)
│   ├── 📁 server2/          # 📡 Node.js + Mediasoup SFU – real-time media streaming
│   ├── 📁 server3/          # 🧠 gRPC service – deep learning inference integration
│   ├── 📄 docker-compose.dev.yml      # Development environment configuration
│   ├── 📄 docker-compose.prod.yml     # Production environment configuration
│   ├── 📄 init.sql                    # Database initialization & permission setup
│   └── 📁 monitoring/                # 📊 Prometheus / Grafana configuration
│
├── 📁 doc/                 # 📚 Documentation and assets
│   └── 📁 images/           # 🖼️ Architecture diagrams, system flow, visuals
│
└── 📄 README.md             # Project overview and setup instructions