The Investigation Process

The investigation team has gathered around the long faux-mahogany table in the field office conference room. At the far end is seated what some call “the injured party”. Personally, I much prefer “the person who got hurt”. Questions are bouncing off the walls, fired in no particular order:  Did you check the clearance? Was there a procedure? What were you doing there? Were you wearing your PPE?
 
It’s like a scene straight out of Twelve Angry Men, if you happen to be familiar with that old movie. The next time you can’t find anything to watch on Netflix, it’s worth checking out if not. Made in the 1950’s, it’s considered a masterpiece by the cinematographers, who, to this day, study the use of camera angles and varying the focal length of the camera lens. Except for its prologue and epilog, it was filmed on a single set: a jury room, with its big wooden table and dozen chairs. 
 
Which explains the title.
 
I first watched Twelve Angry Men in college, for a course on group behavior. The group dynamics part flew right over my head. What I found captivating was how Henry Fonda succeeded in convincing eleven adults – his fellow jurors – to see things his way. One by one. 
 
Reminds you a lot of safety leadership, doesn’t it?
 
The Investigation
 
The opening scene, set around the table, regularly plays out on industrial stages the world over.  Officially it’s known as an investigation. Have the misfortune of a headline-making event or a serious injury, or do significant damage to equipment or the environment, a formal inquiry likely will be commissioned, as per policy. Like a root canal, consider yourself fortunate when you’re not invited to be part of the process.
 
The good news is that high profile events like those are few and far between. But, as you know, found at the other end of the spectrum are the relatively minor events, which are not always so small as to fall outside the requirement of reporting, investigating, and submitting a proper explanation. 
 
In your capacity as a leader, there can be occasions when you find yourself doing the reporting, running the investigation, writing up the findings, or reviewing the final work product. Having spent three decades in the chemical business, as you might expect, I’ve done them all.  Just not necessarily well. That opening scene was based on a true story, as they say in the movies. I was the one asking the question, “Was there a procedure?” 
 
Since we’re into full disclosure, I might as well admit, having been hurt twice in the first two weeks of my industrial career, I’ve also been in the hot seat. Thankfully, I wasn’t hurt that badly, so I was not invited to sit in the Plant Manager’s conference room to explain.
 
Experiences like these shape perceptions, and they can also motivate and focus interest and action. Years later, my solution was to become a champion of using a systematic method to find root cause.
 
I’m sure you’re familiar with both the term and the process. You likely work for an organization that makes use of one of the popular root cause methodologies.
 
How’s that working out for you?

The Root Cause
 
The investigation process isn’t limited to safety, health and environmental events; there are plenty of other failures that get a thorough going-over. But there’s no getting around the simple fact that execution of root cause analysis is limited to failures. If it were otherwise, you’d regularly be reading detailed reports on what went right.
 
No, it doesn’t have to be that way. Yes, it is. Feel free to change the things you can, assuming you want to. Remember, I am just a consultant and writer. 
 
The appeal of employing a formal root cause methodology is that it lends objectivity and removes emotions from the investigation. “Follow the process” was the advice we got when we took our root cause training from one of the experts. Following the methodology we were taught, we started at the end and finished at the beginning.
 
In reality, no matter the problem, where you end up – the “root cause of failure” – can only be one of two reasons: by an Act of God, or action (or inaction) by we humans. The former has been known to happen, whereas, the latter happens all the time. 
 
Since everyone knows that, the predicable work-around has become to employ cause categories like “management system failure” and “design flaw” and “rule known but not followed” and “lack of proper training.”  Pointing the finger of blame at inanimate objects and faceless and nameless people conveniently satisfies the need to know the root cause.
 
Don’t think the technique is limited to finding fault with friends and colleagues. In one high profile case, the National Transportation Safety Board elected to find “the City” as the principal cause as to why a flatbed trailer was driven across a rail crossing – right in front of a speeding train. Take your pick: either it’s a hundred thousand or so residents, or the seventy five square miles of dirt they lived on.
 
By the way, you might find this interesting: the NTSB does not use the term “root cause”. They label their conclusions  as “probable cause.”  That term was imported from criminal law, where it’s the standard to obtain a search warrant.
 
That is a scary thought. 
 
Here’s a different idea: what about just turning the whole thing over to Artificial Intelligence? Let the computer ask all those pesky questions, add up the answers and spit out the cause, root or probable.
 
That way, everyone could get mad at the computer, and the computer could care less.
 
What’s The Point?
 
I will plead guilty to “stirring up the pot”, but this process is too important to “leave well enough alone” as your grandmother would neatly sum up the choices. To the extent you are now thinking about the investigation process in your operation, I have succeeded in making my principal point.
 
Since we are now well into the subject, try answering three simple questions about the investigation process:
 

1. Is “finding the root cause” the point of the process?
2. Is “learning a root cause method” the same as knowing how to properly investigate an event?
3. Is there a standard for what constitutes a “quality investigation”?

 
The questions are easy to ask and simple to answer. However, you might find the answers troubling. I do.
 
Deming Speaks
 
Like Twelve Angry Men, W. Edwards Deming likely came along long before your time. A statistician by avocation, Deming is widely acknowledged as the Father of Statistical Process Control. In the post Second World War era, he took his expertise off to Japan at a time when their label on consumer products was synonymous with junk. Lean Manufacturing and Six Sigma improvement practices are built on the foundation of his once famous Fourteen Points. 
 
Dr. Deming’s mantra was “All work is a process.” A process converts inputs into output; every process can be improved; variability should be reduced.  So, let’s apply a little “Deming thinking” to those three questions (keeping in mind this is a newsletter, not a book.)
 
Is “finding the root cause” the point of the process?
 
Not in the opinion of a process improvement expert like Bill Wilson.  A Six Sigma Black Belt and teacher, his view on investigating production problems is that “Problems are treasures – and the parallel to near-misses.” That isn’t to celebrate the waste afflicting the customer, by rather to shine light on issues previously hidden in the darkness of a factory on the journey for continuous improvement. Effective problem solving is a unique opportunity to define and manage variability in the people, process or technology elements of the process.
 
How useful is knowing the root cause to the goals of “shining light…and continuous improvement”?
 
Is “learning a root cause method” the same as knowing how to properly investigate a problem?
 
From first-hand experience, I can tell you many of the techniques necessary to properly investigate any kind of failure were not part of my extensive root cause methodology training. Knowing how to frame questions, listen to answers, evaluate information, create and test a hypothesis, and develop and evaluate potential solutions are critical competencies for any investigator. 
 
Is there a standard as to what constitutes a “quality investigation”?
 
Before Deming came along, judging quality was an entirely subjective matter. He helped to settle, once and for all time, what defines quality: it is conformance to the requirements. 
 
As to what those requirements are, the choice ultimately rests with the customer. What exactly are the requirements for an investigation? Who determined them? What should they be? 
 
And, first and foremost, who really is the customer?
 
The Last Word
 
I’ll be the first to admit that there’s a lot here to digest. But the investigation process is one of those Moments of High Influence, where followers are paying close and careful attention to what their leader say and do.
 
Recognize that, and you’ll appreciate the investigation process deserves your attention.
 
Paul Balmert
June 2024