AI-powered visual quality control system Rev 360

Overview
AIS Rev 360 is a Deep Learning based 360° inspection solution that removes the need for part orientation when evaluating surface finish on cylindrical packaging. Using an array of side cameras, the system generates precise image unwraps to analyze labels, sleeves, caps and surface defects on glass bottles, PET containers and similar revolution packaging.

This solution is a revolution in the world of 360 unwraps
It provides real-time visualization of the production line and realistic analysis of 3D objects without manual handling. The unwrap quality and processing speed enable fast, accurate inspection on high-throughput lines without reorienting parts.

What Does AIS Rev 360 Do?

• Detects defects in labels, sleeves, caps and surface finish across 100% of production.
• Automatically ejects defective products to maintain uninterrupted production flow.
• Stores production data for traceability and later analysis.
• Supports simplified format parameterization for quick line setup and changeovers.

Certification / Recognition
IDI-20221085_AIS REV – Awarded by the Center for Technological Development and Innovation (C.D.T.I.) on December 22, 2022, for the experimental development of a 360° Deep Learning solution for quality control of cylindrical packaging.

Key Features

• Side-camera configuration that inspects every detail of the product finish without requiring orientation.
• Compatibility with a wide range of revolution packaging types (glass, PET, etc.) for industrial integration.

High-Precision Visual Inspection

• Analyzes product finish details on the full 360° surface without reorienting parts.
• Ensures complete 100% inspection with automatic rejection of non-conforming items.

Capture of Superior Quality Images

• Uniform controlled lighting that highlights all areas of the product.
• High-resolution cameras (≥ 5 Mpx) for sharp image capture.
• Capture speed typically reaching 4–5 containers per second for efficient inspection.

Advanced Technology in Electrical Panel and Mechanical Design

• Rosepetal controller for real-time neural network processing.
• Controlled strobe lighting and proprietary electronics to ensure optimal capture conditions.
• FPGA with patented microcontroller for precise synchronization and easy adaptation to existing lines.

Training and Intelligent Modules

• Pre-trained neural networks to enable defect detection from commissioning.
• Rosepetal Deep Learning software (developed by AIS Vision Systems) with extensive storage and remote access for real-time monitoring.

Comprehensive Quality Control (360°)

• Detects defects on labels, sleeves, caps and other elements to guarantee 100% inspection coverage.
• Effective inspection of transparent packaging and labels, enabling detection on translucent materials.
• Automatic ejection of defective products to preserve production continuity.
• Production data storage for post-run analysis and simplified format parameterization.

Technical specifications

• Cameras: side-camera configuration; high-resolution cameras (≥ 5 Mpx).
• Capture speed: approximately 4–5 containers per second (typical, dependent on configuration).
• Inspection coverage: 360° inspection for cylindrical packaging (labels, sleeves, caps).
• Lighting: uniform controlled lighting; strobe lighting with proprietary electronics.
• Processing: Rosepetal controller for real-time neural network processing; FPGA with patented microcontroller for synchronization and line adaptation.
• Software: pre-trained neural networks; Rosepetal Deep Learning software with remote access and extensive production data storage.
• Functionalities: 100% automated inspection, automatic ejection of defective products, production data storage, simplified format parameterization.
• Compatibility: glass bottles, PET containers and other revolution packaging formats.