Ignivis - AI-Driven Personalized Heat Risk Detection

Problem Statement

Heatwaves and extreme environmental conditions are increasing globally, especially in countries like India. Existing weather applications only provide general temperature or AQI data but fail to assess an individual's real-time physiological stress or personalized heatstroke risk. Outdoor workers, elderly individuals, and vulnerable populations often suffer heat-related illnesses due to lack of personalized and proactive monitoring systems.

There is currently no integrated system that combines computer vision-based physiological analysis with environmental intelligence to predict individual heat stress risk in real time.

Proposed Solution

We propose an AI-based system that detects real-time human heat stress using computer vision and integrates it with environmental data to predict heat exhaustion risk.

The system works in three layers:

1. Physiological Stress Detection (Computer Vision)

   • Uses OpenCV and MediaPipe for facial landmark detection

   • Analyzes facial redness, eye fatigue, blink rate, and skin brightness

   • Generates a Physiological Stress Score

2. Environmental Intelligence

   • Fetches live temperature, humidity, UV index, and AQI using weather APIs

   • Calculates Heat Index and Environmental Stress Score

3. Hybrid AI Risk Engine

   • Combines physiological and environmental scores

   • Predicts Heat Risk Level (Low / Moderate / High)

   • Provides real-time preventive recommendations

Key Features

• Real-time face-based stress detection

• Hybrid ML-based risk prediction model

• Personalized recommendations (hydration, rest timing, exposure control)

• Voice-based alerts for accessibility

• Risk trend tracking over time

Technologies Used

• Python

• OpenCV

• MediaPipe

• PyTorch / TensorFlow

• FastAPI

• Weather API Integration

• Machine Learning & Deep Learning Models

Impact

This system can help:

• Construction workers

• Farmers

• Traffic police

• Elderly individuals

• Outdoor delivery personnel

By predicting heat exhaustion before it becomes critical, the system aims to reduce heatstroke incidents and improve occupational safety.

Uniqueness & Innovation

Unlike traditional weather apps, this solution:

• Detects actual human physiological heat stress using computer vision

• Combines internal body signals with external climate data

• Provides real-time personalized risk assessment

• Works as a preventive AI health assistant