How Computer Vision Ensures Every Ply Is Accounted for in Wind Turbine Blade Manufacturing

Why Ply Layup Accuracy Matters

In wind turbine blade manufacturing, ply layup is a foundational step. Each blade consists of hundreds of composite layers—known as plies—carefully arranged to achieve the necessary strength, stiffness, and fatigue resistance. These plies are typically made of fiberglass or carbon fiber, laid in specific orientations to handle the complex loading conditions experienced by the blade over its 20–30 year life span.

The accuracy of ply placement is essential. A missed or misaligned ply can compromise structural performance and, in worst cases, lead to failure in the field. For example, several failure investigations by blade OEMs and research institutions—including work by Sandia National Labs—have identified missing or misoriented plies as contributing factors in spar cap cracking, delamination, and fatigue damage. Because wind blades are subjected to millions of loading cycles, even a single inconsistency in the layup can reduce fatigue life and increase the risk of unexpected failure.

This risk is especially relevant today as turbine blades continue to increase in size. With offshore blades now exceeding 100 meters in length, ensuring consistent layup quality at scale is both more important and more challenging than ever.

Quality Strategies for Ply Verification

To manage the risk of ply-related defects, blade manufacturers employ a range of quality control strategies—each with trade-offs in terms of effectiveness, labor requirements, and scalability.

Manual Inspection and Ply Workbooks

The most common approach today involves manual verification. Workers follow detailed ply books or digital work instructions that describe each layer's material, size, location, and orientation. As each ply is placed into the mold, the operator marks it as complete—either on a paper checklist or a digital tablet.

This method is simple and easy to implement, but it comes with limitations:

• Labor-intensive: Every layer requires human attention, tracking, and sign-off.
• Error-prone: Fatigue, distraction, or time pressure can lead to plies being missed or logged incorrectly.
• Limited traceability: Handwritten records or simple digital logs often provide little insight into what actually happened on the factory floor.

While manual inspection may work well for low-volume production or smaller blades, it becomes difficult to scale while maintaining consistency in high-throughput environments.

RFID Tagging and Scanning

To improve traceability, some factories have explored RFID-based systems, where each component in a kit is tagged with an RFID chip, and workers scan the tag before placing it in the mold. The system logs the presence of each component, allowing quality teams to monitor progress and maintain a digital audit trail.

However, in the context of wind turbine blade manufacturing, this approach quickly runs into practical limitations:

• Overwhelming scale: A single blade can have 400 or more individual plies. Tagging every single one with RFID chips would create enormous material overhead and cost—not to mention the logistical complexity of managing and applying hundreds of tags per blade.
• On-mold cutting: Many plies are not pre-cut ahead of time; they are cut directly from rolls on the mold. This makes it infeasible to pre-tag all materials before they are placed.
• Dependent on human compliance: Even if tags were applied, operators would still need to remember to scan each one. A missed scan could falsely indicate a missing ply, introducing more complexity rather than reducing it.
• No spatial verification: RFID systems only confirm that a tagged item is present, not where it’s placed or whether it’s properly oriented. For composite layup, spatial accuracy matters as much as presence.
• Hardware challenges: RFID systems introduce additional infrastructure that must function reliably in harsh factory environments, including interference from metal molds and overlapping signal zones.

While RFID can improve traceability in certain manufacturing settings, it is not a practical solution for ply-level tracking in large-scale composite layup. The sheer number of layers, the frequent use of on-mold cutting, and the critical need for spatial verification mean that RFID tagging introduces more overhead and complexity than it solves in this context.

AI-Powered Computer Vision for Passive Ply Validation

An emerging approach involves using overhead cameras and computer vision to automatically monitor ply layup in real time. In this system, cameras mounted above the mold capture images during the layup process. Each ply includes a label—such as a printed ID —and the vision system uses machine learning to detect and log each ply as it’s placed.

Key advantages of this approach include:

• Passive validation: Unlike manual checklists or RFID scanning, no operator action is required. The system identifies and confirms each ply automatically as part of the layup process.
• Increased traceability: Every ply is recorded with a timestamped image, providing clear evidence of its placement.
• Potential for spatial verification: Depending on system configuration, the vision system can be trained to recognize not just the ply ID, but whether it has been placed in the correct region of the mold. Read more about spatial verification here.

This approach is particularly well-suited to large, repetitive manufacturing lines where manual validation is time-consuming and inconsistent. By capturing objective data at the point of layup, computer vision systems can help identify issues earlier, reduce the burden on operators, and improve the overall quality of blade production.

Moving Toward Scalable Quality Control

As turbine blades grow in complexity and manufacturers strive for greater throughput, the limitations of manual and semi-automated layup validation methods are becoming more apparent. Ensuring every ply is placed correctly is fundamental to blade performance, and relying on human memory or checklists alone presents an unnecessary risk.

Computer vision offers a scalable, non-intrusive way to validate ply layup as it happens—capturing ground truth data that improves traceability, accountability, and ultimately product quality. While not a replacement for skilled technicians or detailed engineering standards, vision systems provide a valuable layer of assurance in an increasingly demanding production environment.

At Leverege, we’ve developed WorkWatch, a vision-based quality assurance tool that enables passive ply validation through AI and camera systems. Designed to integrate into existing workflows, WorkWatch helps teams monitor each layer without adding steps for operators, providing both real-time validation and a digital record of the layup process.

As the wind industry continues to push the boundaries of size and performance, improving the reliability and traceability of composite layup is a practical step forward—and one that vision technology is well positioned to support.