Digital Transformation in Steel Inspection
With increases in steel production come increases in steel quality samples , which in turns adds to the demand on inspectors ( ex : longer hours and / or more inspectors ). Thus , several issues arise from the existing inspection process :
Labor Intensive QA Process
Long Inspection Hours
Lack of Consistency
Product Quality Data Fraud
Steel Inspection Process Issues
Because the grading process requires human inspectors to grade each sample , the grading process can quickly become the largest bottleneck in the manufacturing process , especially when the number of veteran inspectors is low .
Steel inspection is sampled per the amount of steel produced and not by a set number of samples per day . As more steel is produced , more samples are required to be graded , which requires longer hours for inspectors .
As inspectors are asked to work longer hours , they can become fatigued , and it becomes increasingly difficult to maintain a level of consistency in the grading process . Increasing the number of inspectors can also create a lack of consistency from one inspector to another .
As the demand for more output increases , there can be pressure to falsify records and report gradings that are not correct . In recent years , it has become public that several steel manufacturers around the world have falsified their quality records , resulting in large regulatory fines , litigation , and damage to company reputation .
While the inspection measurement itself is not a repetitive task , much of the initial setup work is ( i . e ., preparing the samples for measurement ); therefore , automation can be applied to some ( but not all ) of the inspection processes . In the past , improvements to image acquisition and processing relied heavily on hardware improvements , so it was sensible to offer equipment upgrades . Today , however , it is not the local hardware that provides the largest improvements in image analysis ; rather , it is the advancements in image processing algorithms that reduce analysis time and improve accuracy . Therefore , the idea of continuous improvements , offered as-a-service , provide greater benefits to the customer without incurring the negative impacts of hardware upgrades ( i . e ., the associated downtime from upgrading equipment ).
As standards develop in the areas of automated inclusion measuring , the system should be designed to take further advantage of new measurement algorithms , autofocus functions , and
IIC Journal of Innovation 75