SafeHat WorkNet

Easily deployable IoT self-healing mesh network for real-time hazard alerts & tracking for rapid response, construction, and hazard teams by leveraging esp32 nodes and a portable Raspberry Pi network.

Comment

SafeHat WorkNet: Revolutionizing Workplace Safety with an Easily Deployable IoT Mesh Network

Watch the Demo Video GitHub Repository Presentation Slides

The Problem

In hazardous work environments like construction sites, industrial plants, and mines, real-time safety monitoring and rapid emergency response are critical. Traditional safety systems often rely on centralized infrastructure, which can be unreliable, difficult to deploy, and may not provide adequate coverage in complex or dynamic settings. Workers need a solution that provides immediate awareness of hazards, enables quick communication, and facilitates efficient emergency response, regardless of their location or the state of the infrastructure.

Our Solution: SafeHat WorkNet

SafeHat WorkNet is an innovative safety helmet system that addresses these challenges using a decentralized, self-healing mesh network of smart helmets. Each helmet is equipped with an array of sensors and an ESP32 microcontroller, creating a robust and adaptable safety net that:

  • Monitors the Environment: Detects hazardous gases (CO2, etc.), temperature, humidity, and light levels.
  • Detects Worker Status: Identifies falls, impacts, and irregular movements using an accelerometer and gyroscope.
  • Enables Rapid Response: Facilitates location approximation and quick communication through a decentralized mesh network.
  • Operates Independently: Functions even when traditional infrastructure fails or is unavailable, thanks to its self-healing mesh network and portable Raspberry Pi server.

Key Features

  • Self-Healing Mesh Network: Built on ESP32 microcontrollers and the painlessMesh library, our network dynamically adapts to changing conditions, ensuring continuous data flow even if nodes fail or move out of range. A unique root node election algorithm ensures optimal network topology based on real-time RSSI values.
  • Custom Sensor Integration: We developed custom drivers from scratch using esp-idf for all sensors (MPU6050, MQ135, DHT22, BH1750), demonstrating deep hardware and software integration expertise.
  • Real-Time Alerts: The system provides immediate alerts to workers and supervisors through onboard LEDs and a centralized dashboard, as demonstrated in our video demo.
  • Location Approximation: Using RSSI values and the principles of the Shannon-Hartley theorem, SafeHat WorkNet provides an estimate of worker locations, even without GPS.
  • Easy Deployment: Powered by a portable Raspberry Pi server, our system can be quickly deployed in any environment, regardless of existing infrastructure.
  • Custom Communication Protocol: We engineered a lightweight JSON-based protocol for efficient data transmission across the mesh network.

How It Works

  1. Data Acquisition: Sensors on each helmet continuously collect environmental and worker status data.
  2. Mesh Network Communication: Each helmet (node) broadcasts its data (including RSSI) using our custom protocol. The network dynamically elects a "root node" based on RSSI to optimize data flow.
  3. Bridge Node: A designated node with a good connection to the server acts as a bridge, relaying data from the mesh network to the Raspberry Pi.
  4. Server-Side Processing: A Flask-based server on the Raspberry Pi receives and stores data in an SQLite database.
  5. Dashboard Visualization: A real-time dashboard displays sensor readings, node status, event logs, and approximate worker locations.

Technology Stack

  • Hardware: ESP32 Microcontrollers, Raspberry Pi (Server), MPU6050 (Accelerometer/Gyroscope), MQ135 (Gas Sensor), DHT22 (Temperature & Humidity), BH1750 (Light Sensor).
  • Firmware: C++ (PlatformIO), painlessMesh Library, custom sensor drivers (esp-idf).
  • Server: Python, Flask, SQLAlchemy.
  • Dashboard: HTML, CSS, JavaScript, Chart.js.
  • Database: SQLite.

What We Accomplished (Hackathon Achievements)

  • Functional Prototype: Developed a working prototype of the SafeHat WorkNet system, including multiple sensor-equipped helmets, a self-healing mesh network, a Raspberry Pi server, and a real-time dashboard.
  • Custom Sensor Drivers: Successfully implemented custom sensor drivers using esp-idf, showcasing our low-level programming skills.
  • Robust Mesh Network: Created a resilient mesh network that dynamically adapts to node failures and changing environmental conditions.
  • Real-Time Data Visualization: Developed a dashboard that provides a clear and informative overview of worker safety and environmental data.
  • Location Approximation Framework: Laid the groundwork for RSSI-based location approximation, a key feature for future development.
  • Live Demo: Our demo showcases the fall detection, real-time alerts on the dashboard, and the self-healing nature of the mesh network.

Future Enhancements

  • GPS Integration: For precise location tracking.
  • Two-Way Communication: Add voice communication between helmets.
  • Machine Learning: Implement predictive hazard analysis and personalized safety recommendations.
  • Cloud Integration: For remote monitoring and data analytics.
  • Dashboard Location Visualization: Display approximate worker locations on the dashboard.

Team

  • Cesar Magana: Software Engineer - ESP32 Mesh Network & Data Serialization - LinkedIn
  • Brighton Sikarskie: Hardware Engineer - Sensor Integration & Mesh Network Development - LinkedIn
  • Samuel Bush: Software Engineer - Server & Dashboard Development - LinkedIn

Built With

Share this project:

Updates