Crowd Management System using Raspberry Pi, Firebase and ML
🔗View on GitHubIoT solution to enforce social distancing by restricting gatherings at a specific location during Covid-19.
Overview
Developed a complete IoT-based crowd management system leveraging Raspberry Pi with infrared sensors to monitor entry and exit. Data was stored on Firebase for real-time monitoring. Implemented a Crowd Predictor using a Python Flask server to anticipate dense areas, enabling proactive crowd management. Established HTTP communication between Node.js web server and Flask server for seamless integration. Collaborated closely with team members and stakeholders to gather requirements and ensure alignment, meeting project milestones effectively.
System Architecture
Distributed IoT Architecture: Raspberry Pi nodes equipped with IR motion sensors count footfall at entry/exit points. Real-time data streamed to Firebase Realtime Database via WiFi. Flask ML Server analyzes historical data using LSTM/Time-Series models to predict crowd density at future time intervals. Node.js web server fetches Firebase data and Flask predictions via REST APIs. Dashboard visualizes real-time crowd levels, heat maps, and predictions. Alert system triggers notifications when density exceeds threshold. Edge caching on Pi reduces Firebase query load. Horizontal scaling: multiple Pi nodes with UUID-based zone mapping.
Setup & Implementation
Configured Raspberry Pi to collect and transmit data to Firebase. Developed Node.js web server to display and monitor real-time crowd data. Built and trained machine learning model for predicting high-density areas. Integrated Flask server with web interface via HTTP requests for proactive crowd management.