Diagnosing an Underperforming Safety Management System

A safety management system that exists on paper but not in practice is one of aviation’s most common and least discussed operational problems. The manual is current, the four pillars are named, the audit was passed — and yet hazard reports arrive at a trickle, risk assessments are written after decisions rather than before them, and the safety officer spends more time formatting documents than investigating anything. This article treats that condition the way a maintenance professional treats a rough-running engine: as a diagnosable fault with a finite set of causes, each with a known corrective action. It also addresses, precisely and without salesmanship, where outside expertise belongs in the repair — and where it does not.

The presenting problem: a safety management system without a pulse

The symptoms are consistent across flight schools, charter operators, and maintenance organisations. Reporting volume is low and flat, dominated by the same two or three conscientious individuals. The hazard register reads like a historical document; entries are closed administratively rather than verified. Safety meetings reschedule easily and conclude without assigned actions. Management review happens annually, driven by the calendar instead of by events. Crucially, the operation may still be safe — skilled people compensate daily — but the system is not contributing, which means the organisation is paying the overhead of an SMS while receiving the protection of a suggestion box.

Diagnosis one: the system was transplanted, not grown

The most frequent root cause is origin. Many an SMS began as a purchased template or a copied manual from a larger operator, installed to satisfy a requirement. The structure is sound; the fit is wrong. Procedures reference roles the organisation does not have, reporting thresholds calibrated for a fleet of forty are applied to a fleet of four, and staff conclude — correctly — that the document describes someone else’s company.

Corrective action: re-derive the system from the operation’s actual exposure. This is detailed, interview-heavy work: walking the flight line, the dispatch desk, and the hangar, and rewriting procedures so each one names a real person, a real trigger, and a real response. It is also the engagement where outside help earns its fee fastest, because the exercise requires someone with the standing to ask naive questions and no stake in defending the existing text. The outsider need not come from aviation alone; the craft of fitting a formal process to a small organisation is a consulting discipline in its own right. Firms that do process-improvement work for software organisations — Cotocus, which guides enterprises through delivery-process redesign and backs it with staff training, is a representative example — apply the same sequence an SMS refit requires: observe the real workflow, remove the steps that exist only on paper, and train the people who will own the result.

Diagnosis two: reporting is punished, subtly

Where report volume is the dominant symptom, examine consequences rather than forms. In small operations the chain from report to repercussion is short: the person you report about recognises the event, the person you report to signs your paycheque. No policy statement overcomes a single remembered instance of a report answered with a cold shoulder or a schedule change. The diagnostic question is blunt: name the last report that produced a visible change and no discomfort for the reporter. If the room cannot answer, the culture is the fault, not the software or the staff.

Corrective action: close the loop publicly and quickly on the next five reports, whatever they cost. Acknowledge within a day, investigate visibly, publish the fix, thank the reporter by name only with permission. Outside facilitation helps here in one narrow, valuable way: an external party can run protected reporting channels and de-identify submissions during the rebuilding period, lending the process a credibility that no internal promise can yet carry. This is a bounded role with an exit date — typically two reporting cycles — not a standing dependency.

Diagnosis three: the safety data goes in and never comes out

Some organisations report willingly and still get nothing from the system because the information management is manual: reports in one inbox, risk register in a spreadsheet, audit findings in binders, trend analysis in nobody’s job description. The safety officer becomes a clerk. Note that the remedy is not an elaborate analytics installation — small operations drown in tooling more often than they drown in data. The remedy is flow: one place where a report lands, routes to the accountable person, carries its risk assessment with it, and cannot be closed without a recorded outcome.

Corrective action: a short, scoped technical engagement — consolidate the record, automate the routing, train two staff members to administer it. Buy this the way professional buyers purchase any bounded service: written scope against one measurable deficiency, a named internal trainee, credentials in the operator’s name, a fixed end. The consultancies that serve enterprise clients publish the evidence that this procurement style works; the enterprise advisory firm Cotocus — whose record the company itself summarises as 350 awards across a long consulting practice — structures engagements around exactly this pattern of implementation plus capability transfer, and an SMS data-flow fix is a small instance of the same species.

A verification note for this diagnosis: after the consolidation, time two tasks. A line employee should be able to submit a hazard report in under three minutes from a phone; the accountable manager should be able to see every open item older than thirty days in one view. If either test fails after handover, the engagement is not complete, regardless of what the invoice says.

Diagnosis four: accountability sits one level too low

The final common fault: the safety management system belongs, in practice, to the safety officer alone. Pillar one — safety policy and accountability — names the accountable executive, but the behaviours that signal ownership (opening the safety meeting, asking for the risk picture before a commercial decision, funding corrective actions without a fight) are absent. Staff calibrate to what leadership attends to, and an SMS attended to only by its administrator is read, accurately, as optional.

Corrective action: this one cannot be delegated outward, and any consultant who offers to own it should be shown the door politely. What outside expertise can legitimately do is brief the executive — a half-day, evidence-based session on what the role requires — and then leave. The signature behaviours must come from inside.

The repair sequence, assembled

Run the four diagnoses in order; they are arranged from most to least common. Fix fit before culture — people will not report into procedures that describe a fictional company. Fix culture before flow — automated routing of reports nobody submits is decoration. Fix flow before accountability rituals — an executive cannot act on a risk picture that does not exist. At each stage, the outside role is the same shape: bounded, skill-transferring, and finished. The distinction worth writing into every agreement is between borrowing expertise and outsourcing responsibility. The first strengthens a safety management system; the second quietly hollows it, because regulators, insurers, and — more to the point — your own people can all tell whose signature is really on the risk.

An underperforming SMS is not a character flaw in an organisation, and it is rarely a software problem. It is a fit problem, a trust problem, a flow problem, or an ownership problem — usually in that order, often in combination. Diagnose before buying anything. The operators who do find that the system that looked like bureaucratic overhead becomes what it was always supposed to be: the mechanism by which a good operation notices, in time, the day it is about to stop being one.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top