By: Bob Rogers
Manufacturing and production environments can benefit from implementing Lean Manufacturing and using Six Sigma management strategies with Statistical Process Control (SPC). Problem solving methods are essential skills necessary for both Lean and Six Sigma.
Manufacturing and production environments can benefit from implementing Lean Manufacturing and using Six Sigma management strategies with Statistical Process Control (SPC). Problem solving methods are essential skills necessary for both Lean and Six Sigma.
This is a short summary of how to approach problem solving in a production environment where production yield is an issue; this approach can also be applied to solve other business problems.
1) PROBLEM STATEMENT
The first step is to write a detailed "Problem Statement" for the production yield issue.
The Problem Statement needs to be concise so the team working on solving the problem has a clear definition of the problem including factual data. The Problem Statement needs to include the definition of the expected results and data to set a clear goal for the problem solving task.
The data in the Problem Statement may include:
Who, How, When, What, and Where information
Measurements, Statistical, and yield data
Production rates: products/hr, $/product, scrap units, scrap $, first pass yields, rework costs etc.
Consider separating a Problem Statement into two or more statements (and projects) if the root cause for the issues is likely to be different. Subsets of a Problem Statement need to be prioritized if the root cause is likely to be the same.
2) ROOT CAUSE STUDY
Writing the Problem Statement should assist in the process of discovering some of the possible variables that cause the production yield issue. The Root Cause Study needs to be a comprehensive effort to consider these possible variables and to take a step back and brainstorm other variables that may have lead to the yield issue. There are many approaches to determining the actual root causes of the yield issue. An important first step is to evaluate a broad range of possible causes then focus on the most likely root causes.
Team Brainstorming methods can be beneficial especially when combined with developing Cause Effect Diagrams (also known as Ishikawa or fishbone diagrams). Creating the Cause Effect Diagram by Brainstorming: Methods, Machinery, Management, Materials, and Manpower factors, may seem like a diversion from finding the likely cause. This structured thought process often triggers ideas that may have been overlooked. The "Effect" in the Cause and Effect diagram is the yield issue; main branches need to be added to detail the possible production causes. This diagram can also be used to document the causes that can be eliminated.
A typical diagram may look like this:
Each branch needs to be investigated and can include ideas such as for Machine: tooling, setup, settings, environmental factors, maintenance, and wear.
Proving and disproving possible causes can be simple. Sophisticated methods may need to be used such as: SPC (Statistical Process Control), DOE (Design Of Experiments using Taguchi methods), and FMEA (Failure mode and effects analysis). Creating: Run Charts, Pareto charts, Histograms, and Process Capability studies (Cpk) can lead to the key variables causing the yield issues. Often the causes are discovered during the process of setting up these statistical charts, other times the key variables are found after significant data gathering and analysis. JMP makes a powerful software program that provides both Statistical and DOE tools.
Once the key variable that caused the yield issue is discovered, the root causes of the variable(s) need to be determined. Root causes are often caused by a chain of events. Use the "5 Why" methods to investigate deeper into the root cause. The iterative process of asking why numerous times often leads to the real root cause.
3) ROOT CAUSE VERIFICATION
The Likely Root Causes need to be verified by recreating and witnessing the yield issue or by analysis of the statistical data gathered after the causes have been corrected. If more than one cause variable is involved, then DOE methods can be used to prove the variable interactions if the variables are not independent.
4) MISTAKE PROOFING
Root causes can re-appear if they are not addressed thoroughly.
The "5 Why" study often uncovers several process issues out that need to be mistake-proofed. Mistake-Proofing (also known as Poka-Yoke) is a mechanism that prevents a process from being done incorrectly or makes the error obvious. Mistake-Proofing needs to be done at: design, fabrication, manufacturing, assembly, and inspection.
5) LESSONS LEARNED
The root cause and "5-Why" studies often provide a lesson learned that can be leveraged to other company processes. Proactive initiatives should be taken to review any other company process issues that can also be improved by the lessons learned during this problem solving effort.
6) GIVE RECOGNITION
Give recognition and reward the people involved in solving the production yield problem!
Useful WIKI resources on Problem Solving:
Problem statement
Brainstorming
Cause and Effect Diagram : Ishikawa diagram
Histogram
Pareto Chart
Check Sheet
Control Charts
Design of experiments
Taguchi methods
Failure mode and effects analysis
5 Whys
Poka-yoke
