Data Consumer Guide

Data Provider View guide Data Consumer View guide Service Provider View guide

Complete Use Case: Wind Turbine Predictive Maintenance

Discover wind turbine sensor data, transfer it securely between connectors, and analyze it with AI-powered predictive maintenance tools to forecast potential failures.

🔍 Discover

→

🤝 Negotiate

→

📥 Transfer

→

🤖 AI Analysis

→

📊 Results

Step-by-Step Guide

Follow this complete workflow to access and analyze wind turbine data

Prerequisites & Authentication

Set up your access to the ARC Dataspace ecosystem

Register at Auth Portal: https://authportal.dataspace2.imsi.athenarc.gr/ ✅ Create an account if you haven't already
Available Connectors:

Provider Connector: edc.dataspace2.imsi.athenarc.gr Source of wind turbine data

Consumer Connector: edc.dataspace3.imsi.athenarc.gr Your connector for receiving data

Discover Wind Turbine Data

Use semantic search to find available datasets

Open KnowDS Discovery Portal: https://arc.knowds.imsi.athenarc.gr/
Search for wind turbine data:

🔍 "wind turbine"

🔍 "predictive maintenance"

🔍 "sensor data"

🔍 "WTG-01"

KnowDS uses semantic search to find relevant datasets across all connected providers
Review search results:
- Check asset descriptions and metadata
- Verify data format (should be CSV)
- Review temporal coverage and update frequency
- Note the provider connector URL
Identify the asset you need:
Example Asset:
Name: Wind Turbine WTG-01 Sensor Data
Provider: edc.dataspace2.imsi.athenarc.gr
Format: CSV
Description: Time-series data including generator RPM, temperatures, power output
Copy the Asset ID: You'll need this for the transfer request (e.g., wind-turbine-sensor-data-2025)

Request Data Transfer

Negotiate and initiate transfer from provider to your connector

Log into your Consumer Connector: https://edc.dataspace3.imsi.athenarc.gr/
Navigate to "Catalog Browser": Browse available assets from connected providers
Search or filter for the wind turbine dataset Use the Asset ID or name you found in KnowDS
View Asset Details:
- Verify asset metadata
- Review contract policies and terms
- Check data access requirements
Initiate Contract Negotiation:
- Click "Negotiate" or "Request Access".
- Review the proposed contract terms.
- Accept the terms (automated in most cases).
- Wait for provider approval (usually instant for open policies).
Start Data Transfer:
- Once contract is agreed, click "Contracts".
- Select the contract you aggreed on and click, click "Transfer".
- Specify the destination "Datasink", where the transferred data will be transferred.
- Configure any other transfer parameters if needed.
- Confirm and click "Initiate Transfer".
Monitor Transfer Progress: Go to "Transfer History" section to track status

Negotiating...

→ Transferring...

→ ✅ Complete!
Now the data should be transferred in the datasink you requested: Once transfer is complete, download the CSV file to your local machine

💡 Tip: The transfer happens between EDC connectors (dataspace2 → dataspace3), ensuring secure and compliant data exchange

Prepare Data for Analysis

Verify and prepare the wind turbine data

Download the CSV file Save it from your consumer connector to your local machine
Verify data quality:
- Open in spreadsheet software or text editor
- Check for required columns (Generator RPM, temperatures, power outputs)
- Verify no corrupted or missing critical data
- Ensure timestamps are properly formatted
Required columns for ML analysis:
- Generator RPM Max._rolling_mean
- Nacelle Temp. Avg._rolling_mean
- Generator CoolingWater Temp. Avg._rolling_mean
- Production LatestAverage Active Power Gen 0 Avg._rolling_mean
- Production LatestAverage Active Power Gen 1 Avg._rolling_mean
- Production LatestAverage Active Power Gen 2 Avg._rolling_mean

Run AI Predictive Maintenance Analysis

Use the Wind Turbine Predictor to analyze the data

Open Wind Turbine Predictor: https://windturbinepredictor.dataspace1.imsi.athenarc.gr/
Choose your input method:

📁 Upload CSV File
Drag and drop or select your downloaded CSV

OR

🔗 Provide URL
Enter direct link to CSV (if accessible)
Supports: Google Drive, MinIO, S3, Dropbox, Direct URLs
Upload your CSV file:
- Click "📁 Upload File" tab
- Select or drag your wind turbine CSV
- Wait for upload confirmation
Click "🚀 Run Analysis"
Monitor processing pipeline:

🤖 ML Prediction
Analyzing bearing temperatures...

→

📊 XAI Analysis
Generating PDP plots...

→

✅ Complete
Results ready!

Processing time: 30-90 seconds depending on data size

View & Download Results

Analyze predictions and explainability insights

📈 Prediction Results

Predicted bearing temperatures for each data point
Statistics: Mean, Standard Deviation, Min/Max values
Download: Enhanced CSV with predictions column

📊 Explainable AI (XAI) Analysis

Partial Dependence Plots (PDP): 6 plots showing how each feature affects predictions
Feature Importance: Which sensors matter most
Visual Insights: Understand model behavior