Get in touch

Wind Turbine

Case Study

Wireless Turbine Blade Monitoring for the U.S. Air Force – From SBIR to Scalable Aerospace Innovation

Client: Sensatek Propulsion Technology, Inc.
Partners: Constellation Energy, Electric Power Research Institute (EPRI)
Project Title: Demonstration of Wind Turbine Blade Health Monitoring Technology
Sector: Renewable Energy – Wind Power
Duration: 36 Weeks
Technologies Used: VideoMagic™ (machine vision), deep learning, phase-based motion amplification, fractal dimension analysis

Challenge

Wind turbine blade failures lead to catastrophic costs: unexpected downtime, expensive repairs, and lost generation. The challenge was to commercialize a non-invasive, scalable blade health monitoring solution capable of detecting structural anomalies—before they become problems—without interrupting turbine operations.

Innovation

Sensatek deployed its proprietary VideoMagic™ platform, which integrates:

  • Real-time computer vision for crack and erosion detection
  • Machine learning (U-Net CNN) for defect classification
  • Fractal Dimension (FD) & NGFD analysis for structural complexity mapping
  • Phase-Based Motion Amplification (PBMA) for vibratory signature visualization
  • Complex pulsation analytics for blade passing diagnostics

The system captured video and high-speed imagery from turbines across three wind farms. Using advanced image processing and AI algorithms, Sensatek detected, classified, and trended anomalies in both time and frequency domains—with no physical sensors required.

Key Results

5 Independent Diagnostic Approaches validated

Including ML segmentation, frequency tracking, pulsation envelope analysis, and fractal damage detection.

No operational disruption during installation or data collection

Sensors were replaced by high-speed cameras (and later, smartphones), reducing complexity.

Matched proximity sensor data

PBMA-derived modal coordinates were calibrated against legacy proximity probes, confirming equivalent accuracy.

Smartphone-based version validated

Enabled rapid, low-cost inspections with iPhone cameras—perfect for utility operators seeking agility.

Detectable anomalies from real-world defects

Subtle blade root fatigue, pitch misalignment, and leading-edge erosion detected before visual damage appeared.

Utility-endorsed performance

EPRI and Constellation acknowledged the platform’s value in lowering inspection time and cost per turbine.

Applications & Scalability

  • Wind turbine blade inspection (onshore/offshore)
  • Blade root fatigue detection
  • Hub bearing looseness and pitch misalignment diagnosis
  • Structural anomaly trending using FD/NGFD
  • High-resolution ODS and motion amplification for predictive maintenance

Commercial Impact

  • 30–40% reduction in O&M costs projected
  • 10-Year NPV: $10.5 million per utility
  • Avoidance of major failures worth $500K–$1M annually
  • Smartphone deployment reduces CAPEX and enables mass scalability

Commercialization Services Provided

Technical Integration & Prototyping

Adapted VideoMagic for wind industry use cases, including deep learning and edge-based computation

Data Science & Model Training

Built CNN models using a curated dataset of erosion, cracks, and delamination images

Sensor-Free Validation & Calibration

Performed displacement calibration vs proximity probes with strong correlation

Deployment Optimization

Engineered for 4G/LTE/Satellite remote connectivity and real-time AWS dashboards

Smartphone Adaptation

Built a handheld/mobile-friendly version to drive faster field adoption by O&M teams

Ready to Commercialize Next?

If your utility or technology firm is seeking to pilot, deploy, or commercialize a scalable, AI-powered solution for equipment health monitoring—without wiring a single sensor—we can help. Let’s build your case study next.