Skip to content

README.md

Latest commit

 

History

History
149 lines (99 loc) · 5.46 KB

README.md

File metadata and controls

149 lines (99 loc) · 5.46 KB
1️⃣ Problem Statement PS 06: Predictive Hospital Resource & Emergency Load Intelligence System
2️⃣ Project Name Hospital Command Center (MediForecast Pro)
3️⃣ Team Name The Validators
4️⃣ Deployed Link 🔗 Deploy Link
5️⃣ Demo Video ▶️ Watch 2-Minute Demo
6️⃣ PPT Link 📄 View Presentation Deck

Python Streamlit Scikit-Learn PuLP

✅ Project Overview

The Problem: Hospitals currently operate on historical averages, leading to critical staff shortages during sudden surges and wasted resources during quiet periods. Furthermore, during city-wide crises, lack of coordination leads to "Patient Stacking" where one hospital is overwhelmed while another nearby is empty.

Our Solution: The Hospital Command Center is a SaaS-ready platform that acts as a "Digital Twin" for hospital operations. It utilizes:

  1. AI-Driven Forecasting: Random Forest models predict patient influx using weather & calendar data.
  2. Resource Optimization: Linear Programming (PuLP) mathematically minimizes staffing costs while ensuring safety ratios.
  3. City-Wide Load Balancing: Automatically routes ambulances to neighboring facilities when capacity breaches 100%.
  4. Supply Chain Intelligence: Forecasts Oxygen and PPE usage to prevent stockouts.

✅ Key Features

Feature Description
📂 SaaS Data Integration Any hospital can drag-and-drop their historical CSV. The system's Retraining Pipeline instantly builds a custom AI model for them.
🚑 Ambulance Routing An intelligent "Air Traffic Control" for ambulances. If ER capacity > 100%, it provides optimal diversion routes to partner hospitals.
📦 Supply Chain Forecast Predicts consumption of critical resources (Oxygen Cylinders, N95 Masks) based on predicted patient pathology.
📋 Mathematical Rostering Generates the perfect shift schedule (Morning/Evening/Night) to minimize cost while maintaining a 1:5 Nurse-to-Patient ratio.
🚨 Emergency Broadcast "One-Click" dispersion system to alert Dept Heads and recall off-duty staff during Mass Casualty events.

✅ Setup & Installation Instructions

Prerequisites

  • Python 3.8 or higher installed.

Step 1: Clone the Repository

git clone [https://github.com/ByteQuest-2025/GFGBQ-Team-the-validators.git](https://github.com/ByteQuest-2025/GFGBQ-Team-the-validators.git)
cd GFGBQ-Team-the-validators

Step 2: Install Dependencies

pip install -r requirements.txt

Step 3: Initialize the System

Run the data generator script once to create the privacy-compliant "Digital Twin" dataset:

python generate_data.py

*Output: Dataset generated successfully.*

Step 4: Launch the Dashboard

Start the application:

python -m streamlit run app.py

The application will automatically open in your default web browser at http://localhost:8501.

✅ Usage Instructions

  1. View Live Status: The top bar shows the system status and Real-Time Temperature fetched from the API for your location (Pune HQ).
  2. Set Prediction Parameters:
  • Target Date: Select a future date to forecast.
  • Temperature: If the date is today, it uses live sensors. If it's in the future, use the slider to simulate weather conditions (e.g., a heatwave).
  • Mass Casualty Toggle: Switch this ON to simulate a major disaster and see how the system handles a sudden 60% surge.
  1. Run Analysis: Click the 🚀 Analyze & Optimize button.
  2. Review Intelligence Report:
  • Predicted Influx: See the exact number of expected patients.
  • Optimized Roster: Check the generated table for the exact number of staff needed for Morning, Evening, and Night shifts.
  • Bed Map: Visual grid showing occupied vs. free beds in ER, ICU, and General Wards.

👨‍💻 Developed by Team: The Validators

Byte Quest Hackathon 2025