Final Year Project

A.E.G.I.S.

A.E.G.I.S. - Autonomous Elderly Guardian & Intelligent Sensing.
A device-free fall detection system using Wi-Fi CSI and mmWave sensor fusion for privacy-preserving vital monitoring.

aegis.py
1class AEGIS:
2 def detect_fall(self, csi_data):
3 # Wi-Fi CSI Analysis
4 features = self.extract(csi_data)
5 if self.is_fall(features):
6 self.check_vitals()
7 self.alert_caregiver()
8 return True

The Problem

Falls are a leading cause of injury and death among the elderly. Current detection methods have significant limitations.

36M+

Falls per year among elderly globally

684K

Fatal falls annually worldwide

95%

Falls occur in care facilities

50%

Patients forget wearables

Limitations of Current Solutions

Camera Systems

Privacy violations, cannot be used in bathrooms

Wearable Devices

Often forgotten or removed by dementia patients

Pressure Mats

Limited coverage, can be tripped over

Our Solution

A.E.G.I.S. combines Wi-Fi sensing with mmWave radar to provide comprehensive, privacy-preserving elderly monitoring.

Privacy-Preserving

No cameras or images. Only analyzes Wi-Fi signal distortions to detect movement patterns.

Wi-Fi CSI Analysis

Extracts Channel State Information from standard Wi-Fi signals to detect falls.

mmWave Vital Signs

Uses radar sensing to monitor breathing and detect unconsciousness after falls.

Real-Time Alerts

Immediate notification to caregivers when a fall is detected.

Device-Free

No wearables required. Works for patients with dementia who may remove devices.

Continuous Monitoring

Works 24/7 in all lighting conditions, including complete darkness.

How It Works

The system uses a multi-stage pipeline to detect falls and monitor vital signs.

01

Signal Capture

ESP32 extracts CSI from Wi-Fi signals at 100Hz

02

Feature Extraction

Raspberry Pi processes signal patterns in real-time

03

ML Classification

TensorFlow Lite model classifies activity patterns

04

Alert System

Immediate notification with vital sign status

Technology Stack

Built with carefully selected hardware and software for optimal performance.

ESP32-WROOM-32U

Hardware

Raspberry Pi 5

Processing

LD2410 mmWave

Sensor

Python

Backend

C/C++

Firmware

TensorFlow Lite

ML

Development Timeline

Track the progress of A.E.G.I.S. throughout the academic year.

Week 1-2

Hardware Selection & Setup

Week 3-4

CSI Data Collection Pipeline

Week 5-6

ML Model Development

Week 7-8

mmWave Integration

Week 9-10

System Testing & Validation

Week 11-12

Documentation & Presentation

Follow the Development Journey

Read weekly updates, technical deep-dives, and supervisor meeting logs documenting the entire development process.

Read the Blog

The A.E.G.I.S. Project

Privacy-preserving fall detection for elderly care using cutting-edge sensing technology.

Learn more
In Development

Autonomous Elderly Guardian & Intelligent Sensing

A device-free fall detection system that uses Wi-Fi Channel State Information and mmWave radar to monitor elderly patients without compromising their privacy. No cameras, no wearables - just intelligent signal analysis.

Explore the Project

Privacy-First

No cameras or image capture

Wi-Fi CSI

Analyzes signal distortions

mmWave Radar

Vital signs monitoring

ML Powered

Intelligent fall detection

12
Week Timeline
6+
Blog Posts
2
Sensing Modalities
2026
Completion Year