Quality

This uses a statistical technique that is typically used for analysing problems relating to process output quality, like excessive defects, excessive variation or lack of capability to meet customer requirements. It can quickly indicate if your existing process is stable and under control and show the effects of a process change that you have made. It also includes a number of statistical tests which can be useful where, for example, you need to compare the results of using different materials or suppliers for a particular process.

Under the same umbrella are included a number of data analysis tools, including Pareto Charts and Ishikawa (fishbone) diagrams, which are used to quantify effects and assist in finding root causes, respectively.

The output from using this technique would typically aim to be a stable process with a variation or defect rate that is within the customer specified limits.

Typical use would follow the steps shown below.

Step 1

Create a process map to understand the complete process. In a large or complex system, keep the scope down to the immediate area of focus but document any immediate upstream and downstream process inputs and outputs as these may effect what you are analysing. For example, in the process below, the quality of the raw material may be a factor in the quality of the shipped components.

 

Step 2

Create a Control Chart to understand variation, mean and stability of the process output. The type of chart will vary with data type and sample size.

 

Step 3

For Continuous data, create a histogram to understand data distribution, centering and process capability and for Attribute data, create a Pareto chart to analyse different types of defect.

 

Step 4

Create one or more Ishikawa (fishbone) charts or Cause-Effect matrices to determine root causes. Then ask “Why?” 5 times to get to the root cause(s). For more simple problems, just asking “Why?” 5 times is often sufficient, as it is not always necessary to categorise the issues.

 

Step 5

Now that you have the root cause(s) of your problem, you need some solutions to remove them. This is where the “thinking” part of the exercise is needed. Traditionally, the 5 W (plus 2 H) have been used to good effect, these being Who, What, When, Where, Why, How and How Much (there are a lot of “W”s in process improvement), with each being questioned for alternative answers. If ideas are still not forthcoming, you might try some Lateral Thinking to get creative. Hopefully though, after a few cups of coffee, so many possible solutions will evolve that you may need to apply a structured rating for each of them.

Some typical factors to use might be:

  • Feasibility to Implement

  • Benefit to Implement

  • Cost to Implement

  • Time to Implement

  • Risk to Implement

Then apply a 1-5 rating for each, sum and sort the outcomes for each solution and you will have a to-do list for the next Step.

Step 6

Implement your preferred changes and analyse the results, by repeating steps 2-3. A small scale test may be required before full implementation, depending in the impact of the proposed change. If you have reached your goal, put some controls in place to ensure your changes are maintained over time. If things didn’t turn out as expected, start another Plan > Do > Check > Act cycle and find out what went wrong and what to do next.