I heard an interesting article “Medical Detectives: The Last Hope For Families Coping With Rare Diseases” on NPR today about a four year diagnostic process for a rare disease on a pair of brothers. This story introduced the “Undiagnosed Disease Network” and I was amazed about their efforts and successes. It also reminds me of the challenges we face in our own work.
The challenge in diagnosing a disease is like defining a problem during the problem solving process. We don’t want to have dentists pull our teeth every time just because “it hurts”. It is no different in industry; we shouldn’t assume the cause of failure is always the same just because “the product failed the same test.”
One of the worst cases that I can remember is where an engineer assumed a dimensional issue was the cause of problem. His problem definition was “the part failed the assembly process and it is oversized per spec” without studying the actual cause of failure. It led to months of work to regain trust from both customers and suppliers. What was the actual root cause? Misalignment of the press. He assumed that an oversized hole caused the fitting failure and shared this with everyone. However, after investigation, the final problem definition was “the seating pressure was too high regardless of shifts, the faults only occurred on a particular line, and the failures only started after the loading fixture was repaired.” Based on this problem definition, it was very clear this is a mis-alignment or undersized issue due to high seating pressure. There was no way the part is oversized. Months of work would have not been wasted if we took time to define the problem properly.
What is a good problem definition? My tip is using the “5Ws”, they are What, When, Where, Who, and hoW. Checking the “5Ws” would ensure the right elements exist in your problem definition. In addition, the solutions often become clear when you have a clear problem definition. Let’s check our example with the “5Ws”; “the seating pressure was too high (What) regardless of shifts (Who), the faults only occurred on a particular line (Where), and the failures only started after the loading fixture was repaired (When & hoW).”
Circling back to the story, it is great to see technology advanced to the point that we can use the human genome to identify the actual causes. This type of story always gives me hope about human beings. We still have lots of smart people working hard to make our society a little bit better every day.