Automated Defect Detection with SCIIL AI VISION
SCIIL AI VISION is a comprehensive software solution for automated AI-based visual quality inspection in industrial production environments. The system analyzes high-resolution camera images in real time and reliably detects visual defects such as wrinkles, scratches, contamination, streaks, or color deviations – even within tight production cycle times.
A particular focus lies on dynamic defects whose appearance changes depending on material, geometry, tension, or process conditions. This is exactly where traditional image processing systems and smart cameras reach their limits.
Automated Quality Inspection as a Response to Skilled Labor Shortages
Manual visual inspections are labor-intensive, subjective, and difficult to scale. At the same time, quality requirements continue to increase while qualified inspection personnel are becoming increasingly scarce.
SCIIL AI VISION fully automates these inspection processes. The system immediately detects deviations, documents them completely, and can directly interact with downstream processes when required – for example by triggering line stops, initiating rework actions, or providing targeted feedback to production. This reduces inspection costs, stabilizes quality standards, and relieves personnel workload independently of the experience or daily performance of individual operators.
Central AI Instead of Isolated Smart Cameras
SCIIL AI VISION deliberately uses high-resolution standard industrial cameras combined with a central deep-learning architecture – instead of locally operating smart cameras.
The key difference:
The AI learns defect patterns independently of position, orientation, or fixed reference images. This enables the system to detect dynamic defects that may appear differently on each product and cannot be reliably identified using traditional image-to-image comparison.
At the same time, the central architecture enables a scalable system design:
- AI models are trained once and then deployed across multiple lines, stations, or plants
- Local retraining of each individual camera is no longer required
- Standard industrial cameras are significantly more cost-effective than traditional smart cameras
Smart cameras vs. SCIIL AI VISION
Smart cameras typically rely on fixed reference images and local processing. They are sensitive to position changes, product variants, or material variations and quickly reach their technical limits when dealing with dynamic defects.
SCIIL AI VISION instead uses AI-based deep learning with centralized data management. The system remains robust against product variations, permanently stores all images and results. It also enables continuous learning, which assures data-driven process optimization.
Efficient Training and Fast Commissioning
Initial AI training requires only a small dataset. Around 100 images are typically sufficient to start, while training phases usually take less than one week.
New product variants or materials can therefore be learned quickly without interrupting ongoing production, and the central configuration enables parallel deployment across multiple production lines and plants.
More Than Defect Detection: A Complete Vision Platform
SCIIL AI VISION is not an isolated AI engine but a database-based quality application with user interface, administration, and integrated shopfloor control functionalities.
The software centrally manages and controls all hardware components such as cameras, lighting, and triggers. Through interfaces with the line control system and the transfer of product information (part number, variant, serial number), the system automatically selects:
- the appropriate AI model
- variant-specific inspection parameters
- optimized lighting settings
The lighting automatically adapts to the product, material, and variant, ensuring reproducible inspection results – even with frequently changing designs.
Automated Wrinkle Detection for Seat Assembly
SCIIL has extensive project experience in seat production and AI-based visual quality inspection, particularly in wrinkle detection. SCIIL AI VISION was specifically developed for reliable detection of dynamic wrinkles – both before and after critical process steps such as steaming.
The system detects wrinkles independently of material, shape, or orientation and is therefore particularly suitable for seat programs with many product variants and changing production conditions.
Static defects such as scratches, color deviations, or assembly errors can optionally be covered by existing smart camera systems (e.g., Keyence) or fully integrated into SCIIL AI VISION. All results are centrally available for traceability, heatmaps, dashboards, and statistical analysis.
SCIIL VisuSteam
SCIIL VisuSteam combines AI-based wrinkle detection with the **control of steaming robots**.
The system detects wrinkles before the process, transfers optimized parameters such as pressure, direction, and duration to the robot, and automatically verifies the result afterward.
Optionally, a second vision station after steaming enables an automatic auto-learning loop, which is based on real production results. Data processing fits the specified cycle time while reaching optimum utilization of the robot capacities.
SCIIL AI VISION QCX
SCIIL AI VISION QCX enables automated end-of-line quality inspection for both dynamic and static defects.
The system reduces the need for manual inspections and therefore significantly lowers the labor costs. Digitization ensures consistent quality, being independent of the experience or availability of inspection personnel.
Customer-Driven AI Training – Without Vendor Dependency
An integrated AI training and retraining module allows customers to independently train, validate, and release new product variants, materials, or defect patterns.
Fixed rule coding or permanent dependency on the software vendor is eliminated. This supports fast product ramp-ups and sustainable series production.
Traceability, Analysis and Prevention
All images, inspection results, defects, and decisions are stored with full traceability. Based on this data, SCIIL AI VISION provides dashboards, heatmaps, reports, and automatic notifications when unusual trends or defect accumulations occur.
This ensures that quality is not only inspected but actively managed, analyzed, and continuously improved.