Last week, we looked at the over-arching process used for correcting and improving within Six Sigma, DMAIC (Define, Measure, Analyse, Improve, Control). DMAIC and Six Sigma are supported by a huge toolkit of quantitative and qualitative tools to support measurement and analysis. Not all of them need advanced statistics or sophisticated training to use with some benefit.
Any competent manager should be building your own personal toolbox and here are six that can be readily and widely applied.
A simple means to get to the root cause of a problem is to start with a statement of the problem and to ask ‘why?’ Then, starting with your answer, ask ‘why?’ again, and repeat until you can go no further. Now you have a root cause.
Why five? There is no magic to five, but it does seem that you rarely need many more stages and too few steps will usually only take you to an intermediate cause. Five seems to be at the sweet spot for many problems.
Also known as the Ishikawa (after Kaoru Ishikawa) Method, this is another way to help find causes. But its emphasis is on breaking down the multitude of causes to an effect. You represent the outcome (often unwanted) as the head of a fish, and then show/facilitate identification of as many causes as possible, representing each as a fishbone.
Some causes are sub-categories or root causes of another cause, creating ever finer fishbones.
SIPOC analysis looks for the source of a problem or poor performance in one of five places, with the:
The Five Cs or 5C Process
It does not get simpler, conceptually, than this. This will help you stabilise, maintain and improve a process or work environment.
- Clear Out
Get rid of clutter and non-essential assets, materials, processes.
Create a tidy and effective working space:
’a place for everything, and everything in its place’.
- Clean and Check
Keep everything clean and use the cleaning process to spot damage, faults and abnormal conditions.
Ensures that everything conforms to the standards that have been set.
- Custom and Practice
Ensure that everyone knows and follows the rules, and understands what purpose they serve.
Box Plots are a good way to plot data to see the effects of variation. Rather than plot single data points, representing an average, such as the average height data for boys, below…
We can plot the ranges of heights for most boys (70%) with a box and nearly all boys (94%) with the bars. This allows us to see two ranges on one chart. Use the box for the commonly occurring range and the bars for either the whole range or, as here, for all but the extreme outliers, as in the chart below…
Failure Modes and Effects Analysis (FMEA)
Perhaps the most complex and sophisticated tool here, so, in a nutshell, we examine every possible failure mode and assign it a score. Scores over a certain threshold lead the failure to be considered ‘critical’.
The score, or ‘Risk Priority Number’ is given by:
RPN = Severity x Occurrence x Detection
The individual scores for severity (how bad the fault is), Occurrence (how frequently it is likely to occur) and Detection (how hard it is to spot prior to release to customer) are calculated separately according to standard tables. Examples of these tables are on the DMAIC Tools website, here.