Quality Control
Challenge
Quality testing is an important part of a manufacturing production line, but this can present challenges.
When quality control is done by humans (as it often is), it can bottleneck the production process in an impactful way.
Moreover, only a sample of the produced goods is inspected, and it is assumed that this is representative of the entire batch.
Solution
Designed and built a Machine Learning Computer Vision model to identify quality issues based on images of produced goods.
Integrated the solution directly into the production line to detect issues in real time.
Built a dashboard for real-time analysis of quality.
Results
Lower production costs by reducing time required for quality control.
Quality control in real-time enables production managers to react to issues more quickly.
An automated system enables the inspection of every part being produced instead of just a sample, improving the quality overall.