Virtek launches AI-powered Iris AI Composite Inspection system for defect detection

Virtek, a provider of precision laser projection and vision-based inspection systems, has launched the Iris AI Composite Inspection system, an AI-powered vision platform designed to detect defects in composite parts during manufacturing. The system uses machine learning algorithms trained on composite defect datasets to identify anomalies such as delamination, porosity, foreign object debris, and surface irregularities in real time. Virtek positions the Iris AI as a drop-in solution for automated layup and inspection cells, targeting aerospace, defense, and advanced air mobility manufacturers who need faster, more consistent quality assurance without slowing production throughput.

The launch addresses a persistent bottleneck in composite manufacturing: manual inspection remains slow, subjective, and difficult to scale as production volumes increase. While automated fiber placement and tape laying have advanced, post-layup inspection has lagged, creating a quality-control gap that limits throughput and raises rework costs. Virtek's Iris AI competes with other vision-based NDT systems from companies like Vention, Siemens, and emerging startups using laser-ultrasound or thermography, but its focus on AI-driven defect classification and integration with existing layup cells gives it a practical edge for mid-volume production environments. The system's value lies in reducing inspection time per part while improving defect detection consistency, which directly supports the qualification and repeatability demands of aerospace and defense programs.

For composite manufacturers, the practical question is whether the Iris AI's defect library and false-positive rate meet the specific acceptance criteria of their customers, particularly in aerospace where NIST or NADCAP standards apply. Virtek needs to demonstrate that the system can be trained on new defect types without extensive re-engineering, and that its output integrates with existing quality management systems. If the system delivers on its speed and accuracy claims, it could shift inspection from a rate-limiting manual step to an automated inline process, but adoption will depend on how quickly Virtek can build reference installations and validate the AI model against real production data.