Problem Solving – ‘it’s all in the detail’

Overview

For problem solving to be effective a structured approach must be followed, I favour 8D. That said it doesn’t really what your preference is as long as you do the following:

·      Implement a containment action to protect the Customer while you solve the problem.

·      Adequately define and profile the problem so you know exactly what it is you plan to solve and are able to confirm when you have solved it.

·      Identify the root cause(s) and confirm it is the root cause(s).

·      Select and implement a corrective action that solves the root cause.

·      Implement a preventative action that prevents the problem returning.

The Root Cause

In this article I will focus on the root cause and specifically in the context of manual assembly. When it comes to problem solving there are a wealth of additional supporting tools that are available to assist in fact gathering and enabling you to determine the factual data that is relevant to your specific problem. The art of effective problem solving is being able to select the most efficient tool for your given problem.

Over the years I have encountered many situations, the one I am going to describe below I hope will be helpful to you when confronted with a similar situation. In this example there were multiple operators undertaking the same manual assembly task using the same tools, jigs/fixtures and materials. The process output was had a very poor yield which was recorded as a collective result. Initially I split the process output down to the operator level to obtain better quality data. This identified significant variance from one operator to the next in terms of yield. Before I drilled into the reasons for the yield variance, I wanted to be sure of the proximity of the defects being created on the part as there could be multiple defects, each of which need to be solved.

A simple pictorial view of the assembled part with a grid overlaid. Using the collected data, it was then correlated by operator, and defect location. From this a simple process was applied. The technique was new to me, you are never too old to learn new skills, ‘BOB’s and ‘WOW’s. You simply identify the best of best and worst of worst.

It was immediately clear that for each positional defect there was one operator who was particularly poor and another that was particularly good, however it wasn’t always the same operator. Now I had to establish what was different that resulted in such a variance in results. In this instance with the agreement of the operators I decided to film each operator assembling one part. While I knew they would be extra careful while I was filming, it didn’t matter as I was looking for differences in technique that could explain the poor yield and in turn identify the root cause.

On review of the videos I was looking for both the root cause in terms of a technique which would lead to a defect and equally importantly to identify a potential solution to the problem from my WOW operator in terms of what is it the they do to consistently make a good part.

From this simple process both of the above were successfully identified and this information resulted in a process with a yield of less than 50% increasing to in excess of 95% and frequently with zero defects.

I hope that this example will help you in your problem solving.