AVIS

Precision

10 mm positional accuracy for component presence and alignment checks

Setup speed

24 hrs to configure new part families instead of weeks

Inspection time

8 seconds of inspection time versus minutes of manual checking

Reliability

100% offline operation with no internet connection required

Transform your smartphone into a precision inspection system.

Turn everyday mobile devices into powerful quality control tools. Our computer vision models work in sequence to validate assemblies and identify individual components.

Control over your assembly process

◆

Complete digital documentation with each inspection

⚛︎

Consistent accuracy across every shift

✦

AI-driven visual inspection adapts to high-mix, low-volume production

AVIS transforms template configurations into real-time detection parameters, then generates comprehensive inspection reports that identify all assembly issues including foreign objects.

The system detects and flags missing or misplaced components, guiding operators to precise corrections.

AVIS instantly verifies correct assembly and highlights all components that meet template specification.

Real-time inspection overlays provide confidence scores and visual feedback for every step of the process.

AVIS identifies and alerts for foreign objects present in the assembly zone, ensuring safety and compliance.

Assembly Detection

Identifies missing or extra components, incorrect orientation, and positional misalignment. Ensures accurate matching against expected configuration.

Surface Analysis

Detects scratches, dents, paint pinholes, and coating defects. Validates identification labels, serial numbers, and other text-based markings.

System Integration

Connects with ERP, QMS, and PLM systems for streamlined workflows. Every inspection logged with timestamps and audit-ready records.

High-Mix Manufacturing

Validate both positional accuracy and surface condition across a mix of components - screws, knobs, and brackets. Assembly was divided into zones, with each zone managed by its own validation model.

Component Marking

Leverages machine learning ability to validate factory markings. A validation of a pipe is first detected by a defect-based model, after which the factory-applied label is read using text recognition functionality.

Quality Documentation

Generate complete digital criteria with uniform criteria that ensure consistent outcomes, eliminating variability between inspectors. Automatically generates time-stamped evidence and exportable reports.