3 Keys Unlock Finding Root Causes

Solving for root cause is like solving a criminal murder case. The biggest difference is you can't question most of your suspects. Unlike the TV shows that have suspects faces slapped on a board with alibi data and motives. While you may have a fish bone diagram ( Ishikawa ) illustrating causes and effects relations. Solving the tech problem the usual suspects like contamination, bad parts process changes and configuration changes can be sorted out fairly quickly if you had controls in placed. Need a consultant or just some training?

There are 3 key areas that must be explored in finding the root cause for increase failure rate or catastrophic results.

    A. Know your system [ What makes it tick? ]
    B. Know the time line when things started yielding bad results.
    C. Double and Triple check your assumptions ( things changed and you never knew)

Knowing your systems behaviors is critical in both the development phase and production phase too catch as soon as possible when something is going wrong. In the development phase your coming up a learning curve as you go from prototype to production configuration.  Testing the equipment under different environments and with a variety of configurations of hardware and software can make cause and effect consistencey all the more challenging. Keep track of configurations in the pre-production phase. In the production phase changes are not suppose to occur unless approved under most contracts and formal documentation. Yes I know this does not happen many times in the real world, because someone thinks its a minor change affecting nothing:-)

Knowing you time line is essential to backtracking a cause's origins. You should have a monitoring system and problem tracking system in place during the developing phase. You know when things have been changed or updated. Take the time to document changes or it bite you later and cost even more time before launch. In the production phase engineering change process is more formal. [use ISO 9000 series quality management systems manufacturers or get certified yourself if your doing production).  Documented processes and process control is priceless. It will allow you to easily trace back to changes and know the vintage of your parts (when they were made). I know it may cost more, but not being able to solve a production problem in a timely manner can put you out of business.

Verifying assumptions is where the many break through occur when discovering the tough to find root causes. Some may ask if a part or assemble meets its requirements than their should not be any significant unexpected problem. For that answer go back to knowing your system section. If you are already in production a major redesign may be too costly in terms of time and dollars (need to get creative).

    Let's illustrate with an example:  There is a plastic component that keep breaking off in the same general location that sits inside the machine. A first blush it looked like a new shipper was handling your parts rougher than had been tested initially.  Further analysis shows that part breakage was happening at a greater than expected rate before the shipper was added. Part dimension were the same as always but surface shine was dull compared to previous parts built 6 months ago. The supplier claimed he had not change nothing a part met the spec.. What he failed to mention is he had subcontracted the mold to another vendor to save money. This vendor to increase thru-put per hour shorten the mold cooling time with rapid cooling. This process change created higher stress points at the corners on this part making it easier to break under a shock load. Now you must create the best solution for your circumstances.

Finding the right root cause can be a daunting task but oh so rewarding. The investigation will reveal some weakness in design or processes which can be improved.

What challenges have you come across when trying to find a root cause??

---