Ultrasensitive Defect Detection

Automate quality inspection on X-rays to spot defects x10 faster with more accuracy

Steel & Pipes
Aerospace & Defense
Oil & Gas

Deepflow automates defect detection
 by instantly labeling anomalies on industrial X-ray images. Our innovative AI approach accelerates quality inspection while ensuring accuracy at a micrometer scale.

Visual Inspection Platform

Intelligent Defect Detection

Deepflow provides quality operators inspecting X-ray images with a familiar viewer interface where defects are automatically detected and labeled. With unique AI algorithms that continuously learn over time from the operators' decisions, detection quality continuously improves over time and progress can be tracked.

Accelerate Quality Inspection

In industrial radiography, the interpretation of X-ray images to spot defects remains a manual process. For quality operators, the large volume of images to inspect lengthens their control cycles, which increases fatigue, risk of errors and negatively impacts productivity. Deepflow uses AI algorithms to remove the tedious and time-consuming aspect of inspecting for defects on X-rays while ensuring none are missed or wrongly labeled. A faster and more accurate defect inspection process is now possible without compromising on quality. For the quality operator on the shop floor, the intelligent labeling approach eliminates fatigue and potential bias in interpreting anomalies on X-ray images.

Industrialize quality inspection

It takes a few seconds for Deepflow to process industrial X-ray images, detect anomalies and label them for quality operators to review. Our AI algorithms are trained to recognize various types of defects (e.g. shrinkage, porosity, cracks, holes and more) on a wide range of materials (e.g. aluminium, magnesium, titanium, nickel and more) with an accuracy of 200 microns. With more than 90% of the defects being labelled automatically, quality operators can allocate their time and energy on the 10% that require in-depth analysis.

Deepflow in numbers



Quality Inspection



to production