Spiral Science and Technology Inc. has released Roboscope, an augmented reality inspection platform that anchors digital defect markers directly on the surfaces of large composite structures. Built for wind turbine blade quality control and expanding into aerospace, marine, and construction, Roboscope eliminates manual measurement and paper-based defect logging by capturing type, location, dimensions, and photographic evidence in a single AR session.

Using iPhone LiDAR, inspectors place AR markers at the exact physical location of each defect with approximately 5 cm placement accuracy. Every marker records the defect type, spatial coordinates, dimensional data, inspector ID, and timestamp. Sessions are auto-registered via a work task router scan, and completed records integrate directly with ERP and QC systems — eliminating manual data reentry.
Roboscope’s Layup Registration Table maps each AR-anchored defect to the exact composite plies affected at that point on the part. The system resolves AR anchor coordinates against the part’s layup geometry, returns the ordered ply stack (material, fiber orientation, and nominal thickness per layer), and cross-references defect depth against the stack to determine which plies are involved. Structural plies trigger automatic escalation, and the full layer-annotated defect record exports directly to ERP or repair work orders with ply-level detail.
Purpose-trained ML vision models handle defect classification, severity grading, structural anomaly detection, delamination identification, and surface condition scoring. Models are trained on domain-specific composite inspection data and updated continuously from session evidence. Edge cases route automatically to human experts. Roboscope runs as a native iOS application optimized for handheld use.
It is commercially deployed for wind turbine blade manufacturing inspection and available for composite structure applications in aerospace, marine, and construction sectors.
More info: spiral.technology/products/roboscope


















