Why your incident investigations keep landing on "human error"

If every investigation ends with "the worker didn't follow the procedure," you don't have an investigation process. You have a process for confirming what you already assumed.

It's a familiar pattern. An incident happens, an investigation runs, and the finding is that someone made a mistake — they took a shortcut, missed a step, didn't follow the rule. The corrective action is to re-train them, re-issue the procedure, or remind everyone to be more careful. Case closed. And then, months later, something remarkably similar happens again.

The reason it recurs is simple: nothing actually changed. "Human error" isn't a cause you can fix. It's a label you attach when you stop looking too early.

The person closest is rarely the cause

By the time an incident reaches the person at the workface, a lot has usually already gone wrong upstream. The procedure may have been written for conditions that no longer exist. The tools or staffing may have made the safe method impractical. The pressure to keep moving may have been baked into the schedule months earlier. The worker is often the last line of a system that had already failed several times — and blaming the last line leaves all the earlier failures untouched.

This isn't about excusing genuine recklessness, which is real and rare. It's about recognising that most incidents are produced by ordinary people doing ordinary work inside conditions that quietly made the unsafe option the easy one.

You can't re-train your way out of a system that was set up to fail.

What a systems-based investigation looks for

Structured methods like ICAM exist precisely to stop investigations halting at the individual. They push past the immediate actions to the conditions underneath, typically across a few layers: the defences that were absent or failed; the individual and team actions; the task and environmental conditions that shaped those actions; and the organisational factors — decisions about resourcing, design, scheduling and priorities — that set the whole thing up.

Each layer you uncover gives you something you can actually change. An absent defence can be added. A task condition can be re-designed. An organisational pressure can be acknowledged and managed. None of those appear if the investigation stops at "didn't follow the procedure."

Two investigations, same incident

Picture a worker bypassing an isolation step. The blame-based investigation finds the worker breached the procedure, and re-trains them. The systems-based investigation asks why bypassing it was even possible — and finds the isolation took twenty minutes on a job allotted ten, that the workaround had been informally tolerated for a year, and that no one had ever fed that reality back into the procedure. One of those investigations prevents recurrence. The other schedules it.

Why independence helps

It's hard to investigate your own system honestly. Internal investigators work within the same reporting lines, budgets and decisions that may be part of the cause — and there's a natural, understandable pull toward findings that don't implicate the organisation's own choices. An independent investigator can follow the evidence wherever it leads, ask the uncomfortable questions, and produce findings that carry more weight with regulators, clients and boards precisely because they weren't generated in-house.

The point of all of it

An investigation isn't there to assign fault. It's there to make sure the next person doing that work is safer than the last. That only happens if the findings reach the conditions that produced the incident — and if the corrective actions change those conditions, rather than simply restating the rule that was already there.