InInspekt delivers a step change in wind turbine blade inspection. The project integrates an autonomous crawler robot, multimodal sensing, and AI-driven data fusion to objectively detect internal and subsurface damage inside rotor blades. This reduces inspection risk, downtime, and uncertainty while providing standardized, high-quality diagnostic data for condition-based maintenance and extended service life.

Project Kick-off

On 4 February 2026, we hosted the kick-off meeting for the research project InInspekt at EduArt Robotik. The first session focused on getting to know the partners, aligning technical and organizational interfaces, and agreeing on the immediate next steps across work packages.

Participants

  • Jost Wittmann (Julius Maximilians-Universität Würzburg)
  • Julian Luster (JMU, last name tba)
  • Daniel Hein (LaToDa)
  • Lars Osterbrink (LaToDa)
  • Olaf Manger (LaToDa)
  • Silke Körner (KIT)
  • Michael Stramm (BAM)
  • Christian Wendt (EduArt Robotik GmbH)
  • Markus Fenn (EduArt Robotik GmbH)


Objectives

Develop an autonomous, robot-centric inspection system for in-blade operations that provides objective, repeatable, and high-precision detection of internal defects — including delaminations, cracks, bonding failures, and lightning damage — for both onshore and offshore turbines.

Approach

  • Autonomous crawler navigating within blade interiors using LiDAR, IMU, and advanced 3D localization.
  • Multimodal payload: RGB cameras, passive IR thermography for rapid coverage, active thermography for subsurface defects, and high-resolution 3D laser scanning for geometric context.
  • Real-time spatial fusion: Visual/thermographic data mapped directly to 3D blade geometry; two-stage workflow from primary screening to targeted high-precision inspection.
  • AI analytics: Deep-learning-based object detection and 3D point-cloud analysis, both online and offline, for anomaly detection, damage classification, and visualization.
  • Calibration & benchmarking: Field-deployable calibration and a standardized open benchmark dataset for validation and future research.


Benefits for Industry

  • Detection of internal damages not accessible via conventional visual methods.
  • Higher detection quality, including early-stage and subsurface defects.
  • Less downtime through faster, targeted inspections; improved safety by minimizing human entry.
  • Standardized, reproducible results for cross-site comparability and better lifecycle decisions.
  • Reduced lifecycle cost and extended blade lifetime; scalable platform transferable to other inspection tasks.


Consortium & Roles

  • LATODA: AI and data analytics (real-time anomaly detection, classification, sensor fusion, ML pipelines).
  • EduArt Robotik GmbH (Coordinator): Mobile robotic platform — mechanics, locomotion, robustness, power, and ROS-based system integration.
  • BAM: Non-destructive testing and thermography (passive/active), validation, standardization transfer.
  • JMU Würzburg: Robotics research in SLAM, 3D laser scanning, calibration, and data fusion for precise autonomous navigation inside blades.


Funding

InInspekt runs within the Digital GreenTech – Umwelttechnik trifft Robotik framework and is funded by the Bundesministerium für Forschung, Technologie und Raumfahrt.

Acknowledgments

Special thanks to Saskia Ziemann and Silke Körner (KIT) for their outstanding support.