Interactive Art Installation Monitoring with InfluxDB and Grafana

An example of a real-time monitoring system for an interactive art installation that combines physical sensors, data visualization, and responsive audio/lighting experiences.

Table of Contents

1. Guide
2. Use Cases
3. Quick Start
4. System Overview
5. System Architecture
6. Installation & Setup
7. Configuration
8. Data Models

Guide

You can follow the guide in this repo for an in-depth tutorial on how to set up InfluxDB and Grafana in seconds.

Use cases

Data monitoring in interactive installations with InfluxDB can be useful for:

Operational Management

• Real-time System Health:

  • Monitor equipment failures before they impact visitors (sensor failures, speaker network issues)
  • Track power consumption to prevent electrical overloads in remote locations
  • Alert staff when devices go offline or sensors produce anomalous readings

• Predictive Maintenance:

  • Historical data reveals equipment degradation patterns
  • Schedule maintenance during low-visitor periods
  • Reduce unexpected downtime that disrupts the experience
  • Track environmental wear on sensitive electronics exposed to weather

Artistic Enhancement

• Data-Driven Creative Decisions:

  • Understand which environmental conditions produce the most engaging responses
  • Correlate visitor engagement with specific audio/visual combinations
  • Optimize the relationship between sensors and audiovisual realtime generated content

• Performance Optimization:

  • Identify dead zones where sensors aren't detecting visitors effectively
  • Balance audio levels for optimal sound propagation
  • Adjust sensitivity thresholds based on seasonal changes (temperature, weather patterns, etc.)

Visitor Experience Insights

• Engagement Analytics:

  • Track dwell times in different zones to understand visitor preferences
  • Identify peak interaction periods for staffing and maintenance planning
  • Measure how weather conditions affect visitor behavior and system responsiveness
  • Document successful artistic moments for replication

• Safety and Accessibility:

  • Monitor visitor density
  • Ensure emergency systems remain functional

Documentation and Research

• Artistic Documentation:

  • Create a permanent record of how the installation evolves
  • Build a dataset for future impact studies, installations or academic research

• Technical Knowledge Base:

  • Build expertise for scaling to other locations
  • Document what works in specific environmental conditions
  • Create templates for similar installations

Business Value

• Operational Efficiency:

  • Reduce site visits through remote monitoring
  • Optimize energy consumption
  • Demonstrate system reliability to stakeholders and funders
  • Support insurance requirements for complex outdoor installations

• Scalability:

  • Proven monitoring architecture for multiple installations
  • Remote management capabilities for distributed sites
  • Data-driven proposals for expansion or improvements

Quick Start

# Start the complete monitoring system
./start_monitoring.sh

# Or run components individually:
python installation_sim.py --demo                    # Simulator with visual dashboard
python influx_bridge.py --simulator-api --interval 5  # Real-time data bridge
telegraf --config telegraf.conf                     # System metrics collection

System Overview

This system simulates and monitors an installation with:

• 3 zones: entrance_clearing, deep_forest, riverside
• 27 trees with strain gauge sensors
• 15 visitor detection sensors (LiDAR)
• Audio/lighting systems LED fixtures and speakers with light and audio responsive to environmental data
• Real-time InfluxDB monitoring with ~204k daily data points

System Architecture

Physical Deployment

Installation Layout:
┌─────────────────┬─────────────────┬─────────────────┐
│  ENTRANCE       │  DEEP FOREST    │  RIVERSIDE      │
│  CLEARING       │                 │                 │
├─────────────────┼─────────────────┼─────────────────┤
│ 9 trees         │ 9 trees         │ 9 trees         │
│ 5 LiDAR sensors │ 5 LiDAR sensors │ 5 LiDAR sensors │
│ 1 audio speaker │ 1 audio speaker │ 1 audio speaker │
│ 1 LED fixture   │ 1 LED fixture   │ 1 LED fixture   │
│ Weather station │ Weather station │ Weather station │
└─────────────────┴─────────────────┴─────────────────┘
         │                 │                 │
         └─────────────────┼─────────────────┘
                           │
                           │
                     Control PC
                           │
                    InfluxDB Server

Data Flow Architecture

[Sensors + A/V devices] → [Zone Hubs] → [Control PC + Telegraf] → Server w/ InfluxDB → Grafana
                               ↓                    |                      ↓              |
                    (RPi/microcontrollers)          |               (Ubuntu/Docker)       |
                                                    ↓                                     ↓
                              (Windows w/ installation control software)      (Data viz/Monitoring/Alerts)

Component Responsibilities

Zone Hubs (Raspberry Pi or microcontroller)

• Collect data from local sensors
• Pre-process and validate sensor readings
• Forward data via LAN to control center
• Manage local audio/lighting responses

Control PC (Windows/Linux)

• Aggregates data from all zones
• Runs Telegraf for system metrics collection
• Hosts TouchDesigner/Max or other software for real-time audio processing
• Manages overall installation state

InfluxDB Server

• Stores time-series data from all sources
• Provides high-availability data storage
• Enables real-time and historical queries

Installation & Setup

Prerequisites

• Python 3.10 or higher
• Instances of InfluxDB and Grafana (Docker recommended for testing)
• Telegraf

Environment Setup

# Clone repository
git clone [repository-url]
cd InteractiveArtInstallation

# Create and activate virtual environment
python -m venv .
source bin/activate  # Linux/Mac
# or
Scripts\activate     # Windows

# Install dependencies
pip install influxdb-client numpy watchdog

Configuration Files

Create and configure the following files:

InfluxDB Configuration

# influxdb/config.py
INFLUX_URL = "https://your-influx-instance.com"
INFLUX_TOKEN = "your-api-token"
INFLUX_ORG = "your-organization"
BUCKETS = {
    "installations": "artistic_data",
    "system_metrics": "system_health"
}

Telegraf Configuration

# telegraf.conf
[global_tags]
  installation = "forest_interactive"
  location = "primary_site"

[[outputs.influxdb_v2]]
  urls = ["https://your-influx-instance.com"]
  token = "your-api-token"
  organization = "your-organization"
  bucket = "system_metrics"

Configuration

Simulator Configuration

The installation simulator supports multiple configuration modes:

Real-time Mode (Default)

python installation_sim.py
# Updates every 30 seconds with realistic timing

Development Mode

python installation_sim.py --interval 10 --demo
# Fast updates with exaggerated changes for testing

Snapshot Mode

python installation_sim.py --snapshot --output test_data
# Single data capture for validation

Zone-Specific Settings

Each forest zone has unique characteristics:

Entrance Clearing

• Higher visitor traffic (peak 9-18h)
• 8 connected sensors per zone hub
• Better WiFi signal strength (-45 to -75 dBm)
• More responsive lighting due to visibility

Deep Forest

• Lower visitor traffic but longer stay times
• 6 connected sensors per zone hub
• Weaker WiFi signal (-65 to -85 dBm)
• More ambient audio focus

Riverside

• Variable visitor patterns based on weather
• 6 connected sensors per zone hub
• Good WiFi signal (-45 to -75 dBm)
• Water-themed audio and cooler lighting

Data Models

Primary Data Types

Environmental Data

{
  "measurement": "environmental",
  "tags": {
    "zone": "entrance_clearing",
    "measurement_type": "weather"
  },
  "fields": {
    "temperature_c": 22.5,
    "humidity_percent": 65.2,
    "wind_speed_kmh": 8.3,
    "atmospheric_pressure_hpa": 1013.2
  },
  "timestamp": "2025-01-15T10:30:00Z"
}

Tree Biometrics

{
  "measurement": "tree_biometrics",
  "tags": {
    "tree_id": "OAK_001",
    "zone": "deep_forest",
    "species": "oak"
  },
  "fields": {
    "strain_gauge_reading": 0.0023,
    "movement_amplitude_mm": 3.2,
    "natural_frequency_hz": 0.8,
    "health_score": 0.92
  },
  "timestamp": "2025-01-15T10:30:00Z"
}

Visitor Detection

{
  "measurement": "visitor_detection",
  "tags": {
    "sensor_id": "LIDAR_ENT_001",
    "zone": "entrance_clearing"
  },
  "fields": {
    "distance_cm": 150,
    "movement_detected": true,
    "confidence_score": 0.95,
    "duration_seconds": 45
  },
  "timestamp": "2025-01-15T10:30:00Z"
}

See the data_structure.md file for detailed info on the data model.