Artificial Intelligence
AI ANALYSIS FOR NDT
FlawML is a revolutionary automatic NDT data analysis solution based on machine learning. It combines human-level inspection performance with the speed and repeatability of automation. FlawML offers defect recognition, data quality assessment and customized reporting, seamlessly integrated to your workflow.
Key Features
FlawML software provides automatic analysis of NDT data to aid the inspector. Powered by AI, FlawML is performant and suitable for a wide range of tasks.
-
Finds potential indications in NDT data
-
Reports findings intuitively, fit to your workflow
-
Validated for each inspection case
Benefits
• Allows to focus on indicated areas, saving resources for what’s important
• An extra pair of eyes that never gets tired – useful for high workloads
• Automates mundane tasks like data quality assessment and reporting
• Can be integrated into fully automatic inspections
• Provides statistics for more fine-grained production control, without added manual work
Applications
Radiography
-
DR, CR or digitized film
-
Weld inspection: detection, measurement and classification of defects
-
SNR measurement
-
Image quality indicator (IQI) evaluation
-
Lead letter text recognition
Ultrasonic testing
-
Mechanized inspections
-
Conventional, phased array or time-of-flight diffraction. Ask us more for other types.
-
Defect detection and classification
Visual testing
-
Image-based inspections
-
Real time or recorded video
TrueflawBox
FlawML software comes packaged in TrueflawBox, our standalone Edge AI unit
-
Fast analysis times with 32 TOPS of AI performance
-
Connect to local network or directly to a PC with an ethernet cable
-
No internet connection needed, can be used on site
-
Easy to use: drag and drop files to TrueflawBox for analysis
-
Customized installations to fit your workflow
File formats
- Radiography: DICONDE, DICOM, tiff, jpeg
- Ultrasonic testing: Ask us more!
- Visual testing: mpeg, avi, vob
- Eddy current testing: Ask us more!
- All NDT: Integration to custom formats