PM2.5 Ghostbuster

"PM2.5 is like ghost. It exists even if it is invisible."

Making invisible air pollution visible through hacked sensors and community-powered data, in low-internet environments

PM2.5 Ghostbuster transforms €12 IKEA VINDRIKTNING air quality sensors into GPS-enabled mobile pollution detectors that reveal invisible environmental threats.

Like ghost-hunting equipment for microscopic particles, this open-source IoT system combines hardware hacking, real-time mapping, and community science—empowering anyone to track air quality as they move through neighborhoods, identify pollution hotspots, and make atmospheric data tangible. Born from a 2023 artist residency in rural Thailand, the project has generated 12 months of citizen-collected environmental data across four provinces.

A transformed IKEA air quality sensor becomes a mobile environmental monitoring device

📍 Art-Science Collaboration

Art Project

Project Name: PM2.5 Ghostbuster
Medium: Interactive Environmental Monitoring Installation / Community Intervention
Exhibition/Activity: Artist-in-Residence Program & Community Workshop
Development Period: May 24-28, 2023 (Artist Residency) | Active Monitoring: May 2023 - April 2024
Venue: Baan Noorg Collaborative Arts & Culture, Nong Pho, Ratchaburi Province, Thailand
Type: Workshop / Public Intervention / Online Platform
Geographic Coverage: Ratchaburi, Kanchanaburi, Bangkok, Samut Prakan (Thailand), Kassel (Germany)
Web-App: map.thalay.eu (system archived due to funding constraints)

Collaboration

Werapol Bejranonda - Engineering
Baan Noorg Collaborative Arts and Culture - Host Organization

This collaborative art-science project emerged from a five-day artist residency in rural Thailand, where community members built mobile air quality sensors together. The monitoring network operated for six months across four provinces, making invisible air pollution visible to local communities before being archived in November 2023.

🔍 The Experience: Revealing Invisible Threats

Imagine carrying a device no bigger than a smartphone that reveals the invisible—microscopic particles floating in the air around you, potentially harming your health with every breath.

PM2.5 Ghostbuster transforms affordable €12 IKEA air quality sensors into mobile environmental detectives that track pollution as you move through your city, your neighborhood, your daily commute.

How People Used It

🏍️ On the Move Mount the sensor on your motorcycle or carry it while walking. The device continuously measures air quality and plots data on a live map accessible to anyone with internet.

🗺️ Seeing the Invisible Streets glow red with dangerous pollution levels. Neighborhoods breathe clean and green. Patterns emerge showing how pollution moves through communities.

📶 Works Anywhere—Internet or Not Connect the device to your phone's hotspot → data appears on the global map within seconds. No internet? The device remembers everything locally and syncs automatically when connection returns.

Perfect for mobile vendors, ice cream bike-shops, or motorcycle commuters: Collect PM2.5 measurements, timestamps, and GPS coordinates throughout the day while offline. When you return home or pass a public WiFi hotspot, all accumulated data automatically transfers to the server and appears on the public map—revealing pollution patterns across your community without requiring constant connectivity.

Bringing the Mobile PM2.5 Detector to the Local Community

What Made It Powerful

💰 Accessible Technology — €12 IKEA VINDRIKTNING sensor (designed by David Wahl) + GPS tracking + WiFi connectivity

🌍 Environmental Democracy — Professional-grade air quality tracking in the hands of anyone who wants to understand the air they breathe

📊 Community Data Network — Twelve months of active monitoring across Ratchaburi, Kanchanaburi, Bangkok, and Samut Prakan provinces generated rich environmental data that official monitoring stations couldn't provide. Each mobile device acts as a moving sensor, collectively mapping pollution hotspots across entire communities at a fraction of the cost of fixed monitoring stations

🌫️ The Problem: When Air Becomes Invisible Danger

In Ratchaburi Province, Thailand, like many communities worldwide, air pollution is a silent crisis.
PM2.5 particles — so small they penetrate deep into lungs and bloodstreams — remain invisible to human eyes while causing serious health impacts.

Official monitoring stations are few and far between, leaving communities with little understanding of their daily exposure. Installing professional monitoring networks is prohibitively expensive, often costing thousands per station.

PM 2.5 and Fires in Central Thailand – March 2024

🎨 Our Solution: Art Meets Science & Mobile Community Monitoring

We created PM2.5 Ghostbuster to bridge the gap between hard science and public understanding.

👻 The "Ghost" Metaphor — By treating PM2.5 pollution as invisible "ghosts" that need detection, we use art and cultural approaches to make environmental data accessible and actionable.

🤝 Cross-Cultural Collaboration — This project represents a meeting of German engineering expertise and Thai community arts, demonstrating how maker culture and creative technology can empower communities to monitor and respond to environmental challenges on their own terms.

🛠️ Built Together — During the five-day artist residency, community members learned to hack consumer electronics, build sensors, and create their own environmental monitoring tools.

💡 Smart Economics — Instead of expensive fixed monitoring stations, mobile sensors carried by community members (vendors, commuters, delivery riders) map pollution across entire neighborhoods. Offline data collection means anyone can contribute—even without continuous internet access. When devices sync at home or public hotspots, the entire community gains visibility into local air quality patterns.

🔧 The Magic: From €12 IKEA Sensor to Smart Detector

🏠 Step 1: Hacking Consumer Hardware

The magic begins with a clever hack of consumer hardware.

Inside each IKEA VINDRIKTNING sits a professional-grade particle detector (Cubic PM1006) that IKEA uses to power its simple green-yellow-red indicator lights.

We tap into the sensor's internal communication channel—called UART—to access the raw numerical data that the sensor produces every second.

Hardware schematic by Chiara Giardi showing the ESP8266 microcontroller and GPS integration

Step 2: Adding Intelligence

Into the original IKEA housing, we install:

🔌 ESP8266 microcontroller (about the size of a thumb) — the device's brain
📍 GPS module — tracks location as you move
📶 WiFi connectivity — sends data to the cloud or stores it locally

The enhanced internals: ESP8266 brain, GPS eyes, and the original IKEA sensor

The microcontroller:

Reads PM2.5 measurements every second
Combines them with GPS coordinates
Packages everything as structured data
Transmits to cloud server using MQTT protocol (when WiFi available)
Stores locally when offline, syncs automatically later

☁️ Step 3: Cloud Infrastructure (The Network Brain)

A Python-based server receives data from all active devices and:

✅ Validates incoming measurements
💾 Stores in a time-series database (InfluxDB)
🗺️ Generates GeoJSON file every minute
🌐 Feeds real-time web map (Leaflet.js)
📊 Displays 30 days of historical air quality patterns

🚀 Step 4: Smart Features for Mobile Monitoring

Movement Detection 🏃 Only publishes new data after traveling 30+ meters → conserves battery and data

Offline Resilience 💾 Device stores measurements in flash memory when no internet → auto-syncs when connection returns → no pollution data ever lost

Community-Powered Coverage 🏘️ Mobile vendors, ice cream bike-shops, and commuters become environmental sensors → collect data across neighborhoods while offline → sync at home/public WiFi → entire community sees pollution hotspots on shared map → replaces need for expensive fixed monitoring stations

Key Features

🏠 Consumer Hardware Foundation - Built on €12 IKEA VINDRIKTNING sensors, making professional environmental monitoring accessible
📍 GPS-Enabled Mobile Monitoring - Track air quality along routes, revealing pollution patterns across neighborhoods and cities
💾 Intelligent Offline Operation - Automatically stores PM2.5 data, timestamps, and GPS coordinates locally when offline, syncs when connection restores—perfect for mobile vendors and commuters without constant internet
🌐 Real-Time Global Visualization - Live web map shows all measurements from all devices with 30 days of historical data, publicly accessible to entire community
📱 Smartphone Integration - Works with phone mobile hotspots for instant data sharing, or operates fully offline until reaching WiFi
☁️ Professional Cloud Infrastructure - Enterprise-grade Python backend with automated alerts and REST API
🔓 Fully Open Source - Complete hardware modifications, firmware, and server code available for replication
🌍 Community-Powered Network - Mobile sensors carried by community members map pollution hotspots across entire neighborhoods, eliminating need for expensive fixed monitoring stations
💰 Cost-Effective Coverage - One €30 mobile device can cover areas requiring thousands of dollars in traditional fixed sensors

If you need to understand more

What It Does (Value for different audiences)

Imagine carrying a device no bigger than a smartphone that reveals the invisible—microscopic particles floating in the air around you, potentially harming your health with every breath. PM2.5 Ghostbuster transforms affordable IKEA air quality sensors into mobile environmental detectives that track pollution as you move through your city, your neighborhood, your daily commute.

When you mount one of these modified sensors on your motorcycle or carry it while walking, it continuously measures the air quality and plots the data on a live map accessible to anyone with an internet connection. The invisible becomes visible: streets that glow red with dangerous pollution levels, neighborhoods where the air breathes clean and green, patterns that emerge showing how pollution moves through communities.

What makes this system powerful is its accessibility. The heart of each device is a €12 IKEA VINDRIKTNING air quality sensor—designed by David Wahl for home use—that we've carefully enhanced with GPS tracking and WiFi connectivity. When you connect the device to your phone's hotspot, it instantly shares data to a global map. Even when internet isn't available, it remembers everything locally and syncs automatically when connection returns. This isn't just environmental monitoring—it's environmental democracy, putting professional-grade air quality tracking into the hands of anyone who wants to understand the air they breathe.

The Experience (What people encounter)

In Ratchaburi Province, Thailand, like many communities worldwide, air pollution is a silent crisis. PM2.5 particles—so small they penetrate deep into lungs and bloodstreams—remain invisible to human eyes while causing serious health impacts. Official monitoring stations are few and far between, leaving communities with little understanding of their daily exposure.

We created PM2.5 Ghostbuster to bridge the gap between hard science and public understanding. By treating PM2.5 pollution as invisible "ghosts" that need detection, we use art and cultural approaches to make environmental data accessible and actionable. This project represents a collaboration between German engineering expertise and Thai community arts, demonstrating how maker culture and creative technology can empower communities to monitor and respond to environmental challenges on their own terms.

How It Works

The magic begins with a clever hack of consumer hardware. Inside each IKEA VINDRIKTNING sensor sits a professional-grade particle detector (Cubic PM1006) that IKEA uses to power its simple green-yellow-red indicator lights. We've learned to tap into the sensor's internal communication channel—called UART—to access the raw numerical data that the sensor produces every second.

Into the original IKEA housing, we install a tiny ESP8266 microcontroller (about the size of a thumb) and a GPS module. The microcontroller reads the PM2.5 measurements, combines them with GPS coordinates from the location module, and packages everything as structured data. When connected to WiFi—either from a home network or your phone's mobile hotspot—it transmits this data package to a cloud server using a lightweight protocol called MQTT.

The cloud infrastructure acts as the system's memory and brain. A Python-based server receives data from all active devices, validates it, and stores it in a time-series database optimized for environmental measurements. Every minute, the server generates a GeoJSON file—a standardized geographic data format—that feeds a real-time web map built with Leaflet.js. This map becomes the public face of the project: an interactive visualization where anyone can see current and historical air quality patterns spanning 30 days of data.

The system includes intelligent features that make mobile monitoring practical. Movement detection ensures the device only publishes new data when it has traveled a meaningful distance (typically 30 meters), conserving battery and data. When internet connectivity drops—common during mobile operation—the device stores measurements locally in its flash memory and automatically syncs them when connection returns. This resilience means no pollution data is ever lost, even in areas with spotty coverage.

Getting Started

For Community Members & Artists

What You Need:

IKEA VINDRIKTNING Air Quality Sensor (available at IKEA stores, Product ID: 804.982.46)
ESP8266 development board (Wemos D1 Mini or NodeMCU, ~$3-5)
GPS module (Neo-6M, ~$8-10)
USB-C cable and power bank for mobile operation
Basic soldering skills and jumper wires

Building Your Device:

Open the IKEA sensor housing (gentle prying, no damage to internals)
Connect ESP8266 to the sensor's UART pins (detailed wiring diagrams available)
Add GPS module for location tracking
Upload pre-configured firmware via USB
Power on and connect to WiFi using smartphone configuration portal

Using Your Device:

Mount on motorcycle/bicycle or carry in backpack with power bank
Connect device to your phone's WiFi hotspot
Watch your measurements appear on the live map within seconds
Share the map link with your community

Workshop Opportunities: We regularly conduct hands-on building workshops. Contact Baan Noorg Collaborative Arts & Culture for upcoming sessions.

Key Results & Impact

Live Monitoring Network

DIY Air-Quality Workshop: Building a PM2.5 Detector at Baan Noorg community in Ratchaburi Province - May 2023

"Flower of pollution"—radial patterns showing daily PM2.5 variations across monitoring routes

High-density PM2.5 measurements revealing pollution hotspots in Ratchaburi Province

Real-World Applications

🏍️ Commuter Exposure Mapping - Motorcycle riders track their daily pollution exposure while offline, data syncs at home
🛒 Mobile Vendor Monitoring - Ice cream bikes, food carts, and delivery riders collect pollution data across neighborhoods during work hours, contributing to community maps without needing constant internet
🏘️ Neighborhood Coverage - Communities identify local pollution sources and patterns without expensive fixed monitoring stations
📊 Citizen Science Data - Distributed mobile monitoring network provides comprehensive data where official stations don't exist
🎓 Educational Tool - Schools use devices to teach environmental science hands-on
🗣️ Advocacy Evidence - Data supports community conversations with local authorities, showing pollution patterns across entire neighborhoods

Project Credits

Core Collaboration

Werapol Bejranonda (Germany) - Engineering, System Architecture, Hardware Design
Baan Noorg Collaborative Arts and Culture (Thailand) - Host Organization, Community Engagement, Cultural Translation

Key Collaborators

Chiara Giardi - Curation, Project Interpretation, Continuator
Nisa Jewcharoen - Coordination and Community Liaison

Workshop planning with curator Chiara Giardi and engineer Werapol Bejranonda

Local Team & Participants

JiandYin, Ploy, Im, Boss - Workshop facilitation and device testing

Community

Local residents of Ratchaburi Province for participation, testing, and ongoing monitoring

Community workshop at Baan Noorg teaching sensor construction and environmental monitoring

Technical Foundation

This project builds upon Hypfer's esp8266-vindriktning-particle-sensor (Apache-2.0), which pioneered MQTT connectivity for IKEA VINDRIKTNING sensors. Our enhancements add GPS integration, offline storage, intelligent publishing algorithms, and mobile optimization while maintaining the Apache-2.0 license.

Open Source Libraries

TinyGPSPlus (Mikal Hart) - GPS parsing
WiFiManager (tzapu) - Configuration portal
PubSubClient - MQTT communication
Leaflet.js - Interactive mapping
InfluxDB & Paho-MQTT - Backend infrastructure

Hardware Attribution

IKEA VINDRIKTNING - Designed by David Wahl for IKEA (Product ID: 804.982.46) Internal sensor: Cubic PM1006 particulate matter detector

License & Usage

This project is licensed under Apache License 2.0 - see LICENSE for details.

What This Means:

✅ Free to use for personal, educational, or community projects
✅ Modify and adapt the hardware/software for your needs
✅ Use in art installations, exhibitions, or public interventions
✅ Create commercial applications (with proper attribution)
⚖️ Patent protection provided for all contributors
📄 Must preserve copyright notices and license information

We Encourage:

Replicating this system in your community
Teaching workshops using our designs
Contributing improvements back to the project
Documenting your local environmental findings
Sharing your story with us

For Developers

🏗️ Complete System Architecture (v3.0.0)

System Overview: Three Integrated Components

PM2.5 Ghostbuster is a unified environmental monitoring ecosystem consisting of:

📱 Mobile Hardware (Arduino/) - ESP8266-based sensors with GPS
☁️ Cloud Backend (Server/) - Python services with MQTT, InfluxDB, REST API
🌐 Web Visualization (Leaflet/) - Real-time mapping interface

Data Flow Architecture

IKEA VINDRIKTNING → ESP8266 → GPS → Local Storage ⟷ WiFi → MQTT → Cloud → Leaflet Map
     ↓                 ↓        ↓         ↓           ↓       ↓       ↓         ↓
Internal PM2.5 → Process → Location → Cache → Phone/WiFi → Broker → Server → Real-time Web

MQTT Communication Protocol

Topic Structure:

pm25/{deviceID}/air          # Main data stream
pm25/{deviceID}/CMND         # Commands to device
pm25/{deviceID}/LWT          # Last Will Testament (offline status)

Message Format (JSON):

{
  "_type": "location",
  "lat": 13.7563,          // GPS latitude
  "lon": 100.5018,         // GPS longitude
  "pm25": 35,              // PM2.5 reading (μg/m³)
  "tst": 1640995200,       // Unix timestamp
  "tid": "AB",             // Device ID (last 2 chars)
  "t": "v"                 // Type: 'v'=vehicle, 'p'=ping, 'f'=first_fix
}

📱 Arduino Firmware Implementation

Production Implementations

The Arduino/ directory contains two production-ready implementations:

1. Fixed Location Sensors (`picoPM25bFixedCoord/`)

For stationary monitoring without GPS hardware:

Uses predefined coordinates in configuration
Optimized for static monitoring locations
WiFiManager captive portal for easy setup
Lower power consumption (no GPS)

2. Mobile GPS Sensors (`picoPM25v4/`)

For mobile monitoring with location tracking:

Real-time GPS positioning (Neo-6M module)
Movement-based intelligent publishing (30m threshold)
Offline SPIFFS storage with auto-sync
OwnTracks protocol compatibility
Battery-powered mobile operation

Key Firmware Modules

SerialCom.h - VINDRIKTNING UART communication

Parses PM1006 sensor data stream
Handles serial timing and validation
Provides moving average calculations

Types.h - Data structures and state management

Device state tracking
Measurement buffering
Connection status management

pico.h - Configuration header

WiFi credentials
MQTT broker settings
GPS configuration
Publishing thresholds

Offline Resilience

// Automatic offline storage when no internet
if (!WiFi.connected()) {
    store_locally(pm25_data, gps_location, timestamp);
} else {
    upload_to_cloud(data);
    upload_cached_data(); // Sync stored data
}

Movement Detection Algorithm

// Smart publishing based on distance traveled
float distance = TinyGPSPlus::distanceBetween(
    last_lat, last_lon,
    current_lat, current_lon
);

if (distance >= PUBLISH_DISTANCE) {
    serialize(gps, 'v');  // 'v' = vehicle/movement
    client.publish(pubtopic, payload, true);
    update_last_position();
}

🖥️ Server Backend Architecture

Modular Python Services

Configuration Management (config/)

settings.py - Centralized configuration loader
.env.example - Secure environment variable template
Environment-based credentials (no hardcoded secrets)

Core Services (src/services/)

mqtt_service.py - MQTT client with auto-reconnection
influx_service.py - InfluxDB operations with batch processing
geojson_service.py - GeoJSON generation for map visualization
alert_service.py - WHO/EPA-based PM2.5 threshold alerting
api_service.py - Flask REST API endpoints

Data Models (src/models/)

air_quality.py - Measurement validation and transformation

Applications

main_data_collector.py - Primary service (MQTT subscriber + API server)
main_mqtt_logger.py - Debug logging service

Server Setup

cd Server

# Install dependencies
pip install -r requirements.txt

# Configure environment
cp config/.env.example config/.env
# Edit .env with your credentials

# Run main data collector (includes API)
python3 src/main_data_collector.py

# Or use PM2 for process management
pm2 start ecosystem.config.js

Database Schema

InfluxDB 1.x Configuration:

Database: pm25gps
Measurement: air_quality
Fields: PM2.5 values, GPS coordinates, timestamps, device IDs
Retention: 720 hours (30 days)

Query Example:

query = '''
SELECT mean("pm25") as pm25, "lat", "lon", "device_id"
FROM "air_quality"
WHERE time > now() - 720h
GROUP BY time(5m), "device_id"
'''

REST API Endpoints

# Health check
GET /api/v1/health

# Latest readings from all devices
GET /api/v1/readings/latest

# Historical data for specific device
GET /api/v1/readings/device/{device_id}?hours=24

# System statistics
GET /api/v1/stats

Alert System

WHO/EPA PM2.5 Thresholds:

PM25_THRESHOLDS = {
    'good': (0, 12),
    'moderate': (12, 35),
    'unhealthy': (35, 55),
    'very_unhealthy': (55, 150),
    'hazardous': (150, 500)
}

Configurable notifications via SMTP when readings exceed thresholds.

🗺️ Web Visualization Frontend

Leaflet.js Implementation

Features:

Real-time GeoJSON data updates (60-second refresh)
Marker clustering for dense data visualization
Heat map overlays with color coding by PM2.5 levels
Historical data slider (720 hours)
Mobile-responsive design
Full-screen mapping capability
GPS location services for user positioning

Data Source:

// Auto-updating GeoJSON feed
const dataUrl = '/gj/pm25gps.geojson';
setInterval(() => {
    fetch(dataUrl)
        .then(response => response.json())
        .then(data => updateMap(data));
}, 60000); // Update every 60 seconds

Color Coding:

function getColor(pm25) {
    return pm25 > 150 ? '#7e0023' :    // Hazardous
           pm25 > 55  ? '#8f3f97' :    // Very Unhealthy
           pm25 > 35  ? '#ff0000' :    // Unhealthy
           pm25 > 12  ? '#ff7e00' :    // Moderate
                        '#00e400';     // Good
}

🔧 Development & Deployment

Hardware Setup

Required Components:

IKEA VINDRIKTNING (€12)
ESP8266 D1 Mini / NodeMCU ($3-5)
GPS Module Neo-6M ($8-10, optional for mobile version)
Jumper wires, USB-C cable
Power bank for mobile operation

Wiring Connections:

VINDRIKTNING → ESP8266
    REST (UART TX) → D2 (GPIO4)
    5V → 5V
    GND → GND

GPS Module → ESP8266 (picoPM25v4 only)
    TX → D4 (GPIO2)
    RX → D3 (GPIO0)
    VCC → 3.3V
    GND → GND

Software Requirements

Arduino Environment:

Arduino IDE 1.8+ or PlatformIO
ESP8266 board package (3.0+)
Libraries: TinyGPSPlus, WiFiManager, PubSubClient, ArduinoJson

Server Infrastructure:

Python 3.8+
InfluxDB 1.x
MQTT broker (Mosquitto)
Web server (Apache/Nginx)

Deployment Commands

Arduino Upload:

# Configure credentials
cp Arduino/picoPM25v4/pico.h.example Arduino/picoPM25v4/pico.h
# Edit pico.h with WiFi/MQTT settings

# Upload via Arduino IDE or:
arduino-cli compile --fqbn esp8266:esp8266:d1_mini Arduino/picoPM25v4
arduino-cli upload -p /dev/ttyUSB0 --fqbn esp8266:esp8266:d1_mini

Server Deployment:

cd Server

# Automated setup
./scripts/setup.sh

# Manual setup
pip install -r requirements.txt
cp config/.env.example config/.env
# Edit .env file

# Start with PM2
pm2 start ecosystem.config.js
pm2 save
pm2 startup

Web Interface:

# Copy to web root
cp -r Leaflet/* /var/www/html/

# Ensure GeoJSON output path is writable
mkdir -p /var/www/html/gj
chown www-data:www-data /var/www/html/gj

Testing & Debugging

Verify MQTT Connection:

# Subscribe to all device messages
mosquitto_sub -h mqtt.thalay.eu -t "pm25/+/air" -u pm25 -P <password>

Query InfluxDB:

influx -database pm25gps
> SELECT * FROM air_quality ORDER BY time DESC LIMIT 10
> SELECT COUNT(*) FROM air_quality WHERE time > now() - 24h

API Health Check:

curl http://localhost:5000/api/v1/health

📚 Project Structure & Version History

Repository Organization

PM25-Ghostbuster/
├── Arduino/                        # Mobile sensor firmware
│   ├── picoPM25bFixedCoord/       # Fixed location sensors
│   ├── picoPM25v4/                # Mobile GPS sensors
│   └── esp8266-vindriktning-particle-sensor/  # Original Hypfer implementation
├── Server/                         # Backend services
│   ├── config/                    # Configuration management
│   ├── src/                       # Source code
│   │   ├── services/              # Business logic layer
│   │   ├── models/                # Data models
│   │   └── utils/                 # Utilities
│   ├── scripts/                   # Automation tools
│   └── ecosystem.config.js        # PM2 configuration
├── Leaflet/                       # Web visualization
│   ├── index.html                 # Main interface
│   ├── css/                       # Stylesheets
│   └── js/                        # JavaScript
├── API.md                         # API documentation
├── CONFIGURATION.md               # Configuration guide
└── DEPLOYMENT.md                  # Deployment guide

Version History

v3.0.0 (2025) - Historical Integration & Complete Documentation

Complete integration of original 2023 implementation
Unified development narrative from prototype to enterprise
Educational documentation of IoT sensor hacking techniques
Community impact recognition and attribution

v2.1.0 (2025) - Professional Features

Advanced alert system with WHO/EPA standards
REST API for data access and system management
Automation scripts and health monitoring
PM2 process management integration

v2.0.0 (2025) - Modular Architecture

Security improvements (no hardcoded credentials)
Modular service-based architecture
Enhanced error handling and auto-recovery
Professional logging and monitoring

v1.x (2023) - Original Prototype

VINDRIKTNING hacking and UART access
GPS integration for mobile monitoring
MQTT connectivity and real-time mapping
Community workshops and initial deployment

🤝 Contributing & Replication

How to Contribute

We welcome contributions from:

Hardware hackers - Improve sensor modifications and power management
Software developers - Enhance backend services and data processing
Data scientists - Develop analysis tools and visualization improvements
Community organizers - Document deployment experiences and workshop curricula
Artists - Create new ways to visualize and interpret environmental data

Contribution Process:

Fork the repository
Create a feature branch
Make your improvements with clear documentation
Test thoroughly (hardware/software)
Submit pull request with detailed description

Replicating in Your Community

Step 1: Assess Your Context

Identify local air quality concerns
Find community partners (schools, arts spaces, environmental groups)
Estimate budget (~€30-50 per device)

Step 2: Build Pilot Devices

Order components (IKEA sensor + ESP8266 + GPS)
Follow hardware modification guides
Test in controlled environment

Step 3: Infrastructure Setup

Deploy server backend (cloud or local)
Configure MQTT broker and database
Set up public web map

Step 4: Community Engagement

Organize building workshops
Train community members in device use
Establish data sharing protocols
Plan ongoing maintenance

Step 5: Data Action

Share findings with local authorities
Use data for community advocacy
Integrate with existing environmental initiatives
Document and share your story

Support & Resources

Technical Documentation: See DEPLOYMENT.md and CONFIGURATION.md
Hardware Guides: Detailed wiring diagrams in Arduino directories
Community Forum: GitHub Discussions for questions and sharing
Workshop Materials: Contact Baan Noorg for curricula and facilitation guides

Related Projects

This project is part of a broader practice exploring the intersection of art, technology, and community engagement. Explore other works:

The Not-So-Modern Dictionary - Interactive language installation
Code of the Sea - Raspberry Pi performance control system
Heating DJ - Thermal imaging music system

Making the invisible visible, one sensor at a time. 👻

Questions or collaboration? Open an issue or reach out through Baan Noorg

Name		Name	Last commit message	Last commit date
Latest commit History 48 Commits
.claude		.claude
Arduino		Arduino
Leaflet		Leaflet
Server		Server
.gitignore		.gitignore
API.md		API.md
CHANGELOG.md		CHANGELOG.md
CONFIGURATION.md		CONFIGURATION.md
DEPLOYMENT.md		DEPLOYMENT.md
LICENSE		LICENSE
README.md		README.md

Folders and files

Latest commit

History

Repository files navigation

PM2.5 Ghostbuster

📍 Art-Science Collaboration

Art Project

Collaboration

🔍 The Experience: Revealing Invisible Threats

How People Used It

What Made It Powerful

🌫️ The Problem: When Air Becomes Invisible Danger

🎨 Our Solution: Art Meets Science & Mobile Community Monitoring

🔧 The Magic: From €12 IKEA Sensor to Smart Detector

🏠 Step 1: Hacking Consumer Hardware

Step 2: Adding Intelligence

☁️ Step 3: Cloud Infrastructure (The Network Brain)

🚀 Step 4: Smart Features for Mobile Monitoring

Key Features

If you need to understand more

What It Does (Value for different audiences)

The Experience (What people encounter)

How It Works

Getting Started

For Community Members & Artists

Key Results & Impact

Live Monitoring Network

Real-World Applications

Project Credits

Core Collaboration

Key Collaborators

Local Team & Participants

Community

Technical Foundation

Open Source Libraries

Hardware Attribution

License & Usage

For Developers

System Overview: Three Integrated Components

Data Flow Architecture

MQTT Communication Protocol

Production Implementations

1. Fixed Location Sensors (picoPM25bFixedCoord/)

2. Mobile GPS Sensors (picoPM25v4/)

Key Firmware Modules

Offline Resilience

Movement Detection Algorithm

Modular Python Services

Server Setup

Database Schema

REST API Endpoints

Alert System

Leaflet.js Implementation

Hardware Setup

Software Requirements

Deployment Commands

Testing & Debugging

Repository Organization

Version History

How to Contribute

Replicating in Your Community

Support & Resources

Related Projects

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases 3

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

1. Fixed Location Sensors (`picoPM25bFixedCoord/`)

2. Mobile GPS Sensors (`picoPM25v4/`)

Packages