Using Machine Vision to Improve Six Sigma – Rigorous Use of DMAIC Can Really Improve Your Product Quality

​Applying the Six Sigma process can improve manufacturing performance and improve quality.  Six Sigma is a set of statistical tools and data-driven process used for process improvement – whether it is for manufacturing, finance, customer service, or any business process that can benefit from process improvement.  Of course for Keox Technologies, we work mainly in the context of manufacturing and improving our customers’ quality and process control.

There are two main methodologies of Six Sigma – DMAIC and DMADV (for “Define, Measure, Analyze, Improve, and Control“ and “Define, Measure, Analyze, Design, and Verify“, respectively).  Keox Technologies follows a variant that is prudently promoted and trademarked by Sterling Innovations Group Inc., DMAIC-T, with the last T for Transition; transitioning the improved process back to the original team as part of the standard operating procedure.

(Besides the name trademarked by Motorola, the name “Six Sigma” defines how the target should be six standard deviations from the process mean to the established process limits.)

So how is Machine Vision used to improve manufacturing process and achieve Six Sigma?  The two typical phases that Machine Vision is utilized are the Measure and Control phases, though vision systems have also been used to help with other phases.

For example, a laminate layer is applied onto an engineered substrate where the width and thickness of the layer needs to be well controlled.  The laminate also exhibited air pockets which were unacceptable.  Substrate speed, laminate pressure, process temperature, and substrate “parallelism” were Defined as some of the factors that affected the laminate quality.  Machine vision, in conjunction with other sensors (eg. infrared temperature sensors) was used to Measure the process and collecting data for Analysis.  Using ANOVA (Analysis of Variance) tests, the significant factors increased laminate quality and minimize variability were identified to be used for process Improvement.  At that point, machine vision and sensors were incorporated to Control the process and, as a project team, ensured that the new process was Transitioned into the organization as part of the standard operating procedure.