# Root Cause Analysis Takes Practice, Patience, and Perseverance

Posted by on Feb 23, 2020 in Problem Solving | 0 comments

Root cause analysis is a broad term that is used to describe a wide range of approaches, tools, and techniques used to uncover the causes of problems.  It is a structured investigation whose purpose is to identify the true cause of a problem and the actions necessary to eliminate it.

I think people often get confused when problem solving because problems sometimes have more than one cause.  Let me give you an example that will clarify what I mean.

Let’s say your problem is high scrap.  A problem solving team might start by putting together a Pareto chart of the monthly scrap dollars by specific part number.  This Pareto chart may show the following:

Looking at this chart shows part number 170 is the largest scrap contributor, accounting for 49.2% of the total scrap for that month.  However upon further investigation, part number 170 has the highest volume of all the part numbers the company produces since it offers all part number options and is the most popular product requested by customers.

Taking another view of the data by department, the problem solving team could construct another Pareto chart.  This Pareto may look like the following:

This chart shows that scrap occurs in every department with welding contributing to over 26%, riveting at 20.6% sub assembly at 18.4%, and so on.  It’s important to notice that all departments contribute to scrap and that no one department is substantially greater than another.  Going one step further, the problem solving team may use a Pareto chart to break down the problems found in each area.  The scrap chart for welding may look like the following:

This Pareto chart shows the team a completely different picture of what’s actually happening.  It tells them specifically the defects occurring in the welding department.  The largest is spatter at 42.7% of the total defects, next inclusions at 21%, porosity at 13.2%, and so on.  Continually drilling down and breaking it up this way gives the team the ability to investigate each of the problems and determine what the causes of each are specifically and the corrective actions required to eliminate them.

Now let’s take another example where there aren’t multiple causes of the problem, but only one.

Let’s say your customer is complaining about product damage.  You ship your product to multiple customers around the world, but only one customer is complaining, a Chinese customer.  The problems occurs frequently and the customer sends you photos of the damaged product, the date of manufacture, the machine, die number, shift, and operator number of the person that made the product.  Since no other customer has ever complained, the team decides the damage is the result of shipping the product overseas in ocean freighter shipping containers and the material getting damaged during the shipment.  They have even put boxes of material on a simulator and observed the shaking of the product that may be causing the damage.  They have spent thousands of dollars improving the product packaging, but complaints continue on a weekly basis.

Someone outside the original team is assigned to solve the problem.  The first thing they do is to analyze the complaint data.  Putting together Pareto charts of the information given to them, they notice that the vast number of complaints are from 2nd shift and that roughly 48% of the complaints are coming from two operators on 2nd shift.  This person then goes to the shop floor and observes the two operators from a distance.  Within a short period of time, he observes how the operators drop the product into the shipping container from a height of 3 feet or higher.  He then has that specific box of material inspected and finds that dropping the material into the container at that height is causing the damage.

As you can see from these two examples, root cause analysis is in no way a well-defined, singular concept but takes good data and analysis.  It is a process that takes looking at data in different ways and multiple levels sometimes using different tools and techniques to let you see patterns in the data.  It takes practice, patience, and perseverance!