Multi-Channel Product Handling
Parallel channels separate, orient, rotate or stabilize products for repeatable imaging. AICIM configurations reach up to 7,200 capsules per minute; capacity depends on the product and selected model.
Precision Inspection · Confidence in Every ProductEnglish
Home/Technology
Inspection Technology
AI Vision Inspection Technology for Solid Dose Quality Control
Purpose-built optics, AI and pattern recognition work with product handling and rejection for pharmaceutical and nutraceutical quality control. Imaging and inspection settings are matched to each product and machine configuration.
Core Technology
Core Technology
In-house imaging, inspection algorithms and controlled product handling work together for repeatable inspection.
Multi-Angle, Multi-Light-Source Imaging
For the capsule configurations described in our brochure, multiple views and three lighting environments can capture up to 30 images per capsule. Imaging is selected for the required surfaces and defects.
Proprietary Inspection Algorithms
Print Compensation and Glare Removal help address print interference and reflections. Character Inspection evaluates markings, while Photometric Stereo Vision reveals surface detail. The applicable methods depend on the product and configuration.
Controlled Rejection and Defect Reporting
Coordinate defect decisions with product rejection. Available reject images, defect distributions and inspection records help teams review settings and investigate production issues.
In-House Development and IP Ownership
AIACVISION develops its own inspection software, imaging methods and product-handling technology. This supports application-specific configurations and continued development.
U.S. Patent US 12,646,153 B2 covers our capsule seam imaging technology. We continue to file patent applications for our proprietary technologies as part of our ongoing innovation.
US 12,646,153 B2
Imaging coverage, algorithms, record availability and throughput vary by equipment model and project configuration.
01 · Optics & Imaging
Make the defect
visible first.
Different products need different lighting, viewing angles and image acquisition. The optical setup is matched to the product and the defects you need to detect.
01Capsules
Shell, joint and fill
Multi-view imaging for capsule surfaces, with optical configurations suited to opaque, transparent, empty and liquid-filled formats.
- Surface and cap–body joint defects
- Transparent-shell and internal-content inspection
02Tablets
Face, edge and sidewall
Product handling and imaging expose the tablet surfaces needed for inspection across different shapes and coatings.
- Cracks, chips and surface abnormalities
- Edges, sidewalls and coating defects
03Softgels
Surface and contents
Lighting is selected for product shape and transparency, helping reveal defects in the shell and contents.
- Shape and surface inspection
- Shell and fill-related defects
04Blister Packs
Product, package and print
Inspection stations and lighting are configured for the product side, foil side and printed information on a blister pack.
- Products and cavities
- Foil, sealing appearance and printing
02 · AI & Algorithms
Two approaches.
One inspection strategy.
Deep learning and pattern recognition provide complementary tools. The method is chosen for the defect, product variation and inspection requirements.
Deep Learning
Train models with representative images and defect annotations for varied visual appearances and new defect types.
Samples → Annotations → Model
Explore the AI PlatformPattern Recognition
Use defined inspection criteria for measurable features such as dimensions, shape and product appearance.
Image features → Defined criteria → Decision
A product-specific inspection recipeCombine the appropriate methods, settings and reject criteria for your application.
03 · Inspection & Rejection
Turn an image
into a production decision.
Product handling, image acquisition, defect evaluation and rejection need to work together. The inspection sequence is configured for the machine and its position in your line.
Present
Feed and position the product for the required views.
Capture
Synchronize cameras and lighting with product movement.
Friendly Recipe Setup and Management
Set inspection criteria, save product recipes and switch recipes for changeover.
Evaluate
Apply the selected recipe and defect criteria.
Reject
Act on the decision using the configured rejection system.
Built around your line
Standalone inspection, integration with processing equipment, and vision modules for blister packaging or counting lines.
Easy Operation & Efficient Changeover
Easy Operation & Efficient Changeover
Prepare for repeat production with saved recipes, clear inspection feedback and practical cleaning and setup procedures.
Intuitive Controls and Saved Recipes
Review machine status, counts and rejection results through organized operator screens. Save and recall product-specific recipes; English and Spanish interfaces are available on supported configurations.
Practical Recipe Setup
Configure Pattern Recognition settings and assign trained AI models to a recipe. Use acceptable products and known defects to evaluate criteria and understand rejection results.
Product Changeover and Cleaning
Accessible product-handling and product-contact components support changeover, cleaning and routine checks. Fit applicable change parts, select the recipe and follow the selected equipment procedures.
Verification Before Production
After initial setup or changeover, run representative samples to verify handling and inspection. Review detection and false rejects before routine production begins.
04 · Production Software
Control the inspection.
Understand the results.
AIACVISION’s in-house control and inspection software supports recipe management, production monitoring and inspection records.
Product Recipes
Store product-specific inspection settings and switch recipes during changeover.
Inspection Review
Review defect images and inspection statistics to understand production results.
User Access & Audit Trail
Control user permissions and record relevant operations.
Reports & Data Export
Keep inspection records and export reports for production review.
Develop models for new products and defects
Use the AIACVISION Online AI Deep Learning Platform for annotation, training, validation and deployment.
Electronic Records and System Integration
Electronic Records and System Integration
Connect inspection information with controlled operation and your production workflow.
21 CFR Part 11 Support
Available functions include individual user accounts, role-based permissions, time-stamped audit trails, batch and inspection records, export and session controls. Electronic signatures, retention, backup and record protection are reviewed for the agreed configuration.
MES and Other System Integration
Define the exchange of product and batch identifiers, inspection counts, rejection summaries, status and alarms. Recipe coordination, start/stop signals, level signals and interlocks are engineered for the agreed project scope.
21 CFR Part 11 implementation depends on system configuration, validation, procedures and customer administration. MES protocols, data fields, transfer direction and interruption handling are confirmed with customer engineering, IT and quality teams. A software version or configuration supporting 21 CFR Part 11 is available as an optional purchase and is provided when selected by the customer.
Start with your samples
See what works
for your product.
Share your product, target defects and inspection requirements. Sample evaluation helps determine the imaging setup, inspection criteria and equipment configuration.
Product formatTarget defectsRequired capacityLine integration
Friendly Recipe Setup and Management
Product-specific recipes connect imaging, defect criteria and rejection decisions. Evaluate changes using representative products and defect samples before production use.
Integration and Configuration
Machine configuration and inspection coverage depend on the product and process. Counting modules, full-surface inspection machines and partial-surface sorters have different roles.
Questions About AI Vision Inspection Technology for Solid Dose Quality Control
How does the system capture product surfaces?
Imaging coverage depends on the machine. AISIM inspects both faces and the sidewall; BASM-12 offers partial-surface screening; counting modules inspect camera-visible surfaces.
How do recipes adapt to product changes?
Recipes match product appearance and defect criteria. Review imaging and handling conditions when the product changes.
How are inspected and rejected products handled?
Inspection results drive the configured rejection workflow. Full machines and counting modules have different handling and rejection arrangements.
Which proprietary inspection methods are available?
AIACVISION technology includes Print Compensation, Glare Removal, Character Inspection and Photometric Stereo Vision. Applicable imaging and algorithms are selected for the product and machine configuration.
Does the software support 21 CFR Part 11 requirements?
Available account controls, role-based permissions, audit trails and electronic inspection records support project-specific implementation of applicable requirements. Configuration, validation, procedures and customer administration determine the implementation.
Can inspection equipment integrate with MES?
Yes. Product and batch data, counts, rejection summaries, status and selected control signals can be engineered for the agreed project scope. Protocols, data fields, interruption handling and interface testing are confirmed with customer engineering, IT and quality teams.
Let's discuss your application
Let’s evaluate
your inspection requirements.
Tell us your product format, target defects and required capacity. We will help you evaluate the next step.