No extra bottleneck
Non-intrusive inline inspection when production continuity matters as much as detection accuracy.
Curf Industry
Inline defect detection for production lines and industrial reality.
Curf helps manufacturers validate and deploy machine vision without turning inspection into a production bottleneck. Our approach combines Synigen, optical validation and non-intrusive inline concepts for complex industrial environments.
Inline first
Designed around existing lines.
Synthetic-to-real
Simulation, defect injection and real optical validation in one engineering flow.
Deploy & Integrate
Runs on real hardware, connects to PLC/MES/ERP, exports KPIs and audit trails.
Why it matters
Industrial AI only sells when it feels safe to implement.
The biggest challenge is rarely initial interest. It is the jump from curiosity to a decision that operations, quality and management can approve. Curf reduces that friction by focusing on one line, one defect family and one measurable next step.
Optics before hype
Camera, lens and lighting studies are validated on real pieces and defects, not only in slideware or model demos.
Decision-ready scope
From object recognition to positioning logic, every step is shaped around what a plant can actually validate and approve.
Core offer
Three building blocks for deployable industrial inspection.
Synigen simulation
Generate realistic training and validation data from CAD geometry, configurable defect types and production-like optics.
Optical validation
Test the right camera, lens and lighting combinations for the real material, finish and defect families on your line.
Inline integration concept
Translate AI output into a workable inline setup with object recognition, positioning logic and production-fit inspection architecture on blanc or painted surfaces.
Focused use cases
From production lines to laser profile verification.
The project work showed clear value in two concrete routes: complete inline defect detection, and laser-based profile inspection for defect detection on micron scale.
Inline AOI system
Designed for object recognition, defect combination over the full surface and realistic deployment constraints.
LPS solution
Working hardware with automatic laser head positioning and validated defect detection.
Safe first step
Start with one production line, one defect family and one validation plan instead of a vague full rollout discussion.
What customers need
A first step that de-risks the whole program.
Assess the line
Define the defect family, line constraints, target KPIs and available data or samples.
Validate the optics and AI
Use simulation and real validation to check detectability, positioning logic and practical inspection setup.
Decide on pilot or rollout
Move forward with a clearer business case, lower integration risk and a more precise implementation scope.
Next conversation
Need an inline inspection concept that operations can actually approve?
We can start from one production pain point and work toward a decision-ready feasibility or pilot scope.