This series has traced one investigation backward, from a fall at a processing plant through six failed barriers, five preconditions, and four management system findings, to a specific confusion about which document — a Risk Assessment or a Job Hazard Analysis — was supposed to govern a task-level decision. This final article turns the direction around: looking forward, at what a properly built Bowtie diagram would have done differently, before any of this happened.
The Same Confusion, Structurally Prevented
The Risk Assessment versus Job Hazard Analysis ambiguity covered in the previous article exists because the two documents can, in practice, be treated as covering the same ground at different levels of detail — leaving room for a team to assume one has already handled what the other was meant to specify. A Bowtie diagram doesn’t have the same room for that assumption, because of how it’s built.
A Bowtie can be constructed at whatever zoom level a job requires — a generic Bowtie covering an activity across an entire site, or a task-specific Bowtie built around one particular operation. The difference is that a Bowtie states its zoom level explicitly, in the Hazard and Top Event that define the diagram itself. There’s no ambiguity about whether a given diagram represents the org-wide picture or one specific task, because the scope is declared as the first thing on the page, not assumed from context or inferred from which document a team happens to be looking at.
Barriers Made Visible, Not Buried in Text
The second structural advantage a Bowtie offers relates directly to the generic-control-measure problem covered earlier in this series. A control measure that reads “wear appropriate PPE, follow SOP, use proper tools” for every task on a spreadsheet-based Job Hazard Analysis looks, on the page, identical to a control measure that’s actually specific and well thought through — they’re both just text in a cell, and vague phrasing doesn’t visually distinguish itself from precise phrasing.
On a Bowtie diagram, the same vague control measure sitting where a specific barrier belongs looks exactly as thin as it is. A barrier on a Bowtie has to occupy a defined position between a threat and a top event, has to be evaluated against whether it’s genuinely specific and independent, and a placeholder phrase sitting in that position is visually obvious in a way the equivalent line in a spreadsheet simply isn’t. The diagram format itself makes it harder for an under-specified control to hide.
What This Would Have Looked Like Applied to This Case
Had the task in this investigation been mapped as a task-specific Bowtie before work began — Hazard: work at height during coupling installation; Top Event: loss of stable body position — the diagram would have forced an explicit answer to exactly the questions that went unanswered in practice. What are the preventive barriers between the threat of an unstable working position and the loss of control itself? Is “maintain three points of contact” specified as a standalone, verifiable barrier, or left implicit in a generic PPE line? Is a working platform listed as a barrier with a defined trigger for when it’s required, or assumed to be unnecessary because the task was labelled “light work” elsewhere in the paperwork?
None of these questions require an incident to surface once a Bowtie forces them onto the diagram. They surface during the diagram’s construction, because a Bowtie can’t be completed without naming each barrier explicitly and positioning it against a specific threat pathway.
Neither Tool Replaces the Other
None of this suggests a Bowtie makes a Job Hazard Analysis or a Risk Assessment unnecessary — each tool serves a distinct purpose across the planning-to-investigation timeline, and Bowtie’s own value depends on the same underlying hazard identification work a good Risk Assessment provides. What a properly scoped Bowtie adds is a structural check against exactly the two failure modes this investigation surfaced: ambiguity about which document governs which decision, and vague control measures that pass review because nothing about their format exposes how little they actually specify.
Closing the Series
Across ten articles, this series has traced one investigation from a single fall to four separate management system findings — none of them pointing at the technician’s judgment on the day. The causation-chain work of Tripod Beta explains, after the fact, exactly why each barrier failed and what allowed those conditions to exist. Bowtie is the tool built to catch the same gaps before an incident forces the question. Used together — Bowtie to design and verify barriers prospectively, Tripod Beta to investigate thoroughly when something still gets through — they cover both directions a genuine safety programme actually needs.
The Practical Question
For the highest-risk routine task at your site, ask whether it’s ever been mapped as its own task-specific Bowtie — not covered implicitly inside a broader Risk Assessment, but diagrammed explicitly, with its own declared scope and its own named barriers. If the answer is no, that’s the same gap this investigation found, sitting quietly until a task specific enough exposes it.
Want to see what a task-specific Bowtie looks like for a job like this? Cikgu Barrier’s Barrier Management: Bowtie Analysis programme teaches teams to build rigorous, correctly-scoped Bowtie diagrams — declared zoom level, specific and independent barriers, and full threat-to-consequence mapping. Available in-house and as a public workshop across Malaysia.