Menu 844-575-9210

Design of Experiments (DOE)

The process no longer generates optimal output, but it used to.  Design of Experiments is the solution.  On day one inputs were optimized but all mechanical processes degrade over time and input values have not changed since the process started.  You know the output is no longer optimal so what do you do to compensate for process degradation?

When one rests on work performed in the past the status quo is entrenched.  The desire to change is nonexistent.  Therefore, the cost of operation is increasing gradually as the process degrades and the control variables are no longer optimized.

Lean Business Position

Lean Business recommends reviewing Run Charts or your other ways of measuring process performance for output degradation.

Next Steps

Select a method to determine optimal settings for optimal output.  The first Design of Experiments methodology was simple and strait forward.  It’s called One-Factor-At-A-Time.  The benefit of this method is that it is straight forward but the disadvantage is that secondary and tertiary effects are ignored.  For tertiary effects, ignoring them is reasonable, but it is dangerous to ignore secondary effects.  Actually a strong secondary effect can add or subtract to a primary factor and give an undetected erroneous result. This will not happen with a Full Factorial Design.

A Full Factorial Design (Full Factorial Design example) allows measurement of all primary, secondary and tertiary impacts of Factors. The advantage is that it is the most thorough methodology.  The disadvantage is that number of experimental runs increases exponentially and the impact of every tertiary effect is measured.  If an experiment is expensive or time consuming then the downside of this method becomes overwhelming.  The best of both worlds is the Fractional Factorial Design.

The Fractional Factorial Design (Fractional Factorial Design example) because it right sizes the number of experiments.  Tertiary terms are seldom or never of a magnitude that will measurably impact the output.  The Full Factorial Design makes the effort to determine these impacts, the Fractional Factorial Design excludes these non-impacts. The disadvantage of the Fractional Factorial Design is its complexity.  A sophisticated Lean Six Sigma Black Belt overcomes this disadvantage.

The Result

Performing a Design of Experiments will bring inputs back into their optimal states, give you profound knowledge of your process, minimize production costs and set up inputs for the future.

« Return

Transform Your Business (844-575-9210) Email
Our Expertise
Build Your Lean Production System Kata Continuous Improvement Velocity: About 1 idea per person per month; Direction: towards vision or challenge Manufacturing Improvement Business Improvement
Government and Military Improvement Hospital Improvement Healthcare Improvement Profound Knowledge
Operations Management System and Process Improvement Methodologies Lean Six Sigma Tools Lean Six Sigma Statistics