The Iceberg Investigation: Why Most Incident Investigations Never Leave the Surface

Incident Investigation and the Iceberg Effect

When was the last time you read an incident investigation report that concluded with:

“Operator failed to follow procedure”

” JSA not followed” or “Contract worker did not follow PTW ”

…”Crew was standing in the wrong place”

Every single time I read this, it’s a cringe moment.

Yes, the operator didn’t follow the procedure. Yes, the crew member was standing in a hazardous position.

But labeling them as root causes and placing blame on the individual? That’s like looking at an iceberg and only documenting what’s visible above the waterline. This ignores the massive structure hidden beneath.

What’s needed is a proper, detailed investigation that unearths the hidden reasons, the systemic failures beneath the surface.

Systemic failures are the deeper, hidden problems – issues with training programs, organizational structure, communication breakdowns, outdated procedures, equipment design flaws, competing operational pressures, and resource allocation decisions.

These are the 90% beneath the waterline that actually determine whether incidents occur.

Why We Stay at the Surface

I’ve reviewed a lot of investigation reports :: by Operations and those cascaded by authorities. The pattern is unmistakable: we seem to often end the investigation based on the tip of the iceberg, we document it thoroughly, blaming someone and then declare the investigation complete.

Here’s the uncomfortable truth about why this keeps happening:

When we blame someone, we feel like we’ve solved the problem.

We disciplined the worker or better we fired em. We retrained the team. We sent out a safety alert reminding everyone to “follow procedures” and “stay alert.” The investigation is closed. The paperwork is filed. Management has no accountability.

Until the exact same incident happens again six months later.

Because we never examined what was beneath the surface. The funny thing is it happens again and again

Let Me Show You What I Mean

Here’s a real LTI scenario I’ve seen repeated across multiple offshore installations. The details change, but the pattern is always the same:

THE INCIDENT: Technician suffers severe hand injury when caught between pipe flange and wrench during valve maintenance.

What Everyone Sees (The Tip of the Iceberg):

• Worker’s hand was in the pinch point
• Worker was sent onshore for surgery to fix broken pinky
• Incident was downplayed to an MTC not an LTI
• Root Cause :: Worker didn’t use the correct tool
• Worker rushed the task
• PPE worn incorrectly

The Investigation Report Concludes: Root Cause: Technician failed to follow safe work practices. Corrective Actions: Retrain on hand safety and proper tool use. Counsel employee on taking time to work safely. Send safety alert to all personnel.

Case closed, right?

Wrong.

What Nobody Investigates (The Hidden Mass Below):

• Maintenance was scheduled during 12-hour night shift when fatigue peaks
• The “correct” tool was locked in a workshop 3 decks away; “wrong” tool was right there
• Production pressure meant 2-hour maintenance window before process restart was required
• Flange design created an awkward body position that made using the proper tool nearly impossible
• The standard procedure was written for a completely different valve configuration
• This was the technician’s first time on this particular valve, despite having 10 years of experience
• Offshore downplays the incident and hence investigation results not scrutinised
• The OIM and Offshore leadership performance KPIs were heavily weighted toward uptime metrics
• Maintenance backlog meant “routine” jobs got compressed into unrealistic timeframes
• The training simulator didn’t include this valve type
• Similar near-misses were reported twice in the past 1 year and only the worker was reprimanded due to the incident investigation

Now look at those two lists again.

Which one reveals the real problem?

Which one gives you corrective actions that will actually prevent the next injury?

The tip of the iceberg is where we assign blame and close the case.

The underwater mass is where we find the solutions that actually work.

Can AI Help Us Dive Deeper?

Here’s where it gets interesting.

A pattern emerges. The language used in blame-focused investigations is remarkably consistent:

• “Failed to…”
• “Did not follow…”
• “Should have known…”
• “Was not paying attention…”
• “Didn’t use proper…”

These phrases flag investigations that stopped at the waterline. They indicate an investigator who found a person to blame rather than a system to fix.

What if we could train an AI model to recognize these patterns? What if AI could detect when your investigation is superficial – still floating at the surface, never reaching the hidden iceberg beneath?

Introducing Our New AI-Powered Incident Investigation Module

At ehs tools, we’re launching a new module that does exactly this.

Our AI-powered Incident Investigation Module is embedded directly into your investigation workflow. As your team analyzes an incident and documents findings, the AI works in real-time to detect whether you’re conducting a superficial investigation or actually reaching the systemic failures beneath the surface.

The AI actively alerts you when:

• Your root cause findings focus on individual blame rather than systemic failures
• You’ve stopped at “operator error” without exploring organizational factors
• Your investigation language indicates surface-level analysis
• You’re missing barrier analysis or defense-in-depth examination
• Your corrective actions target people instead of systems
• Similar systemic patterns appeared in past incidents but aren’t being addressed

Think of it as a persistent alert system that won’t let your investigation team stay at the waterline. When your draft findings say “technician failed to verify valve position,” the AI doesn’t just accept it – it flags the investigation as superficial and prompts:

Investigation Alert: Your findings appear to focus on individual actions. Have you identified the systemic factors that made this error possible? The hidden iceberg may still be unexplored.

The Bottom Line

The Titanic didn’t sink because of what was visible. Your safety program won’t either – unless you have the courage to look beneath the surface.

Or better yet – unless you have AI that won’t let you stay at the surface.

At ehs tools, we are launching an extension to our new AI-Powered Incident Investigation Module – built around a simple principle:

Your investigation isn’t complete until you’ve reached the hidden iceberg.

Our AI analyzes your investigation in real-time and actively alerts you when your findings are superficial – when you’re still floating at the waterline instead of diving to the systemic failures beneath.

If your incident investigations keep finding “operator error” but your incident rates aren’t improving – you need a brave, independent AI that can tell you your investigation is superficial.