An old certificate, no Level 1 finding for years, an operations manual kept up to date: the operator feels safe until the email arrives announcing an unscheduled visit on short notice, on the sole subject of compliance monitoring. Nothing has happened. Several ramp inspections carried out abroad have found, within a few weeks and on different aircraft, the same minor documentation discrepancy. It is not the discrepancy that triggers the visit, it is its repetition, and the fact that no one internally saw it coming. The scenario is banal, and it sums up what continuing oversight actually measures.
What issuance certifies, and what it does not
An air operator certificate attests that, on a given date, before a team of inspectors, an organisation demonstrated that it had the means, procedures and people required for the operation applied for. It is a snapshot. It says nothing about that organisation six months later, and both systems draw the same conclusion: the certificate is issued without a term, and this unlimited lifespan is matched by oversight that never stops.
The document itself is brief. The real content of the authorisation lies in the specifications annexed to it: the types and registrations that may be operated, the areas and nature of operations, the special approvals — low visibility, reduced vertical separation, extended diversion time, dangerous goods, performance-based navigation. It is this annex that moves, not the certificate. An operator almost never loses its certificate in one stroke: it watches it shrink, line by line, as approvals are frozen or withdrawn. Sudden loss exists; erosion is the rule.
From product control to system control
During initial certification, the authority controls a product: manuals, training programmes, a continuing airworthiness management organisation, demonstration flights, and the process ends in a binary decision. After issuance, the object of control changes in nature. The authority audits the auditor. The authority stops sampling the product to sample the system that produces it: it checks that the operator detects its own deviations, analyses their cause and closes them out. In other words, it audits the auditor.
That is the role of the compliance monitoring function, required on the European side with independence from the activities it audits, and it is the role of the assurance component of the safety management system on the American side. The practical consequence is counter-intuitive: an operator whose internal audits find nothing worries an inspector more than one that regularly reports findings. The inspector assumes an audit function that is not actually looking, and will check by reopening one of the domains declared compliant.
For the practitioner, the rule fits in one sentence: a deviation found by your own system, tracked, analysed and closed with evidence, is an asset in the oversight file. The same deviation found by the authority is a liability, and it counts against you.
After certification, regulators don’t just check your paperwork—they check if you’re checking yourself. If your own audits find problems and you fix them properly, that’s good. If they find nothing, regulators worry you might be missing real issues. The key is to catch and solve problems before they catch you.
An airline operating Airbus A320s on 2,500m runways in 30°C conditions finds the same minor logbook error on three aircraft during ramp checks. The authority doesn’t care about the error itself but sees it as proof the airline’s internal quality control system isn’t working. The airline gets an unscheduled visit, not for the error, but for failing to detect it three times in a row.
The oversight cycle is not a fixed calendar
The European model builds, for each operator, an oversight programme that covers all applicable domains and closes on a planning cycle whose reference duration is on the order of two years. This cycle is not a vested right: it can be extended when demonstrated safety performance justifies it, and it shortens as soon as the risk profile deteriorates. It is fed by the complexity of the operation, the history of findings, the flow of occurrence reports, ramp inspections and changes underway.
The effect is cumulative and asymmetric. A clean year lengthens the interval slowly; a Level 1 finding shortens it immediately. Oversight is not spread evenly either: it concentrates where the system has already shown a weakness, and a finding on flight time management will reopen that domain in the following cycle, whatever the quality of the rest. This is what makes the strategy of « preparing for an audit » illusory: what actually gets prepared is the previous cycle.
On the American side, a team assigned to the certificate
The federal architecture does not proceed by audit campaigns but by assignment: each certificate holder is matched with a certificate management office and a team of principal inspectors — operations, maintenance, avionics. The relationship is continuous and named: the operator does not re-explain its organisation at every visit, but it also does not start from a blank page after a bad episode.
The assurance system that structures this oversight breaks the operator down into elements — operational control, crew training, airworthiness monitoring, load handling, each special approval — and questions each one from two angles: design, meaning the adequacy of the written procedure to the requirement, and performance, meaning its actual application. An impeccable procedure poorly applied and a correct practice with no written procedure therefore do not produce the same finding, even though they look alike on the ground.
The underlying logic is the same as in Europe; the unit of measurement differs: an audit plan by domain on one side, a continuous flow of inspections by element on the other. One divergence has everyday effects. On the American side, the detail of the authorisation lives in the operations specifications, amended by the certificate management office: every change goes through the authority. On the European side, a comparable share lives in the operations manual, whose amendments follow the change-management regime. The second system delegates more and bills that autonomy in documentary rigour.
| Subject | European system | US federal system |
|---|---|---|
| Certificate issued | Air operator certificate, without term as long as compliance is maintained | Air carrier certificate, without term under the same conditions |
| Where the detail lives | Annexed operations specifications and approved operations manual | Operations specifications, amended directly by the certificate management office |
| Who oversees | Authority of the operator’s state, teams formed by audit campaign | Certificate management office, principal inspectors named and assigned |
| Pace | Oversight programme closed on a planning cycle, modulated by risk profile | Continuous flow of inspections, priority set by data already collected |
| Unit examined | Regulatory domain | System element, judged separately for design and performance |
| Finding classification | Level 1 and Level 2, distinct timeframes and consequences | Sorted by intent and recurrence: compliance action or enforcement path |
| Possible consequences | Limitation, suspension or revocation, in whole or in part | Civil penalty, action against the certificate, emergency order |
| Disclosure route | Occurrence reports and the just-culture principle | Voluntary disclosure programme, subject to strict conditions |
| Who oversees the authority | Standardisation inspections by the European agency, in addition to ICAO oversight audits | Internal audits of the federal administration, in addition to ICAO oversight audits |
What triggers an unscheduled visit
Scheduled oversight is only half the story. The other half is reactive, triggered by signals that, taken in isolation, look harmless. The most misunderstood is repetition: three identical minor discrepancies on three different aircraft are not worth three times one minor discrepancy, they are worth a failure of the function that should have detected them — whereas each one, taken alone, would have been closed within a week.
Ramp inspections provide the most visible illustration. A finding raised at a foreign stopover does not stay there: it is forwarded to the authority of the operator’s state, which adds it to the file and compares it with previous ones. Treating a ramp inspection as a local incident is a mistake of scale: the measurement is of the operator, through the aircraft that happened to be there.
| Signal | What the authority reads into it | Usual response |
|---|---|---|
| Same minor discrepancy found on several aircraft in a short time | The internal detection function is not working | Targeted visit on compliance monitoring |
| Previous findings closed late, or reopened | Root-cause treatment is superficial | Shortened cycle and review of the domain |
| Designated post vacant or extended interim | The approved organisation no longer exists as such | Request for approval, then a finding if the situation persists |
| Rapid growth of the fleet or network | Same people, doubled workload, no resizing | Audit of the management system and resources |
| Wet-leased aircraft | Operational responsibility diluted along a chain | Review of the contract and of supplier oversight |
| Occurrence report flow collapses | The reporting channel has closed | Interview on just culture, review of the system |
| Report from a staff member | Gap between documentation and actual practice | Unannounced visit, with protection of the reporter |
The finding, its classification, and what it entails
The European system classifies deviations into two levels, and the classification determines everything else. A Level 1 finding characterises a significant non-compliance that lowers the safety level or seriously endangers flight safety: it calls for immediate action by the authority, which may limit, suspend or revoke the certificate, in whole or in part. « In part » is the most frequent form. A Level 2 finding opens a bounded timeframe, during which the operator must get a corrective action plan accepted, address its root cause and prove closure.
The federal system reasons first on intent and recurrence. A deviation that is unintentional, not repeated, committed by an organisation that has the capacity and the will to correct itself falls under a compliance action: root cause, plan, verification. A deliberate deviation, one that recurs after correction, or one from an organisation that refuses to correct itself, moves into the enforcement path — civil penalty, action against the certificate, emergency order. There is also a route Europe does not formalise to the same degree, voluntary disclosure, which protects an operator that reports its own deviation, stops it immediately and corrects it under an accepted plan.
The trap of the corrective action plan. A finding closed with « reminder of the instruction to the staff concerned » and a signed acknowledgement does not work twice. The second occurrence is no longer read as a deviation: it becomes proof that the organisation does not know how to analyse a cause, which shifts the finding from the operational domain to the management system — and that is where classifications turn severe. A variant of the same trap: an extension request filed on the day of the deadline, which proves the timeframe was not used.
Change, where certificates are lost quietly
Most challenges to a certificate do not begin with an incident: they begin with a poorly classified change. Both systems distinguish what requires prior approval from what the operator implements itself, under a change-management procedure that is itself approved. Without it, every significant change goes back to the authority and the company suffers a delay at every evolution; with it, the company gains autonomy and inherits traceability that an audit will check line by line.
Positions of responsibility are the most sensitive point. The accountable manager and the nominated persons — flight operations, ground operations, training, continuing airworthiness — are not boxes on an org chart: the authority accepted them by name, on the basis of their real authority over the resources. An unreplaced departure, an interim that settles in, a post-holder who combines two positions incompatible with the workload: these are natural candidates for a severe finding, without a single flight having been affected, and they are visible from the outside on a simple org chart.
Then come large-scale changes: a new type, a new base, a new area, a wet or dry lease, rapid fleet growth, a change of control. Each one adds to the workload of the same nominated persons, and that is the inspector’s question: is the approved organisation still sized for what it does?
The reflex that avoids it. Keep a change register in which each line carries four columns: the nature of the change, the clause of the approved procedure that authorises it, the trace of the notification or approval, and the date of actual implementation. It is reread at management review and answers the first question of any audit — what has changed since last time? Producing it in ten minutes buys credit that nothing else provides.
The other side: the economic licence
An impeccable technical certificate does not guarantee the right to sell a seat. The European system separates the certificate, which attests to technical capacity, from the operating licence, which attests to economic capacity: financial fitness, insurance, ownership and effective control. The licence presupposes a valid certificate, the reverse is not true, and it can be suspended for reasons where the word safety never appears once. The American system makes the same separation by entrusting it to two administrations: the civil aviation authority for safety, the department of transportation for economic fitness and nationality.
This separation does not mean the two sides ignore each other. An inspector who sees signals of financial strain build up — turnover among management staff, maintenance deferrals pushed to the limits of the programme, spare-parts stocks thinned out, training cut back to the strict regulatory minimum — does not wait for failure: it tightens oversight. Cash-flow pressure counts as a leading indicator. because it shows up in the reverse order of visibility, first where no one is looking.
What to go check in your own documentation
Three checks give, in half a day, a fairly accurate picture of an operator’s exposure. The first concerns the operations specifications: compare each authorisation with what is actually practised. An approval that is kept but no longer exercised — therefore no longer trained nor checked — is a finding waiting for its inspector. The second concerns the change-management procedure: take the last three significant changes and check that they followed the path it describes, with the corresponding trace. The third concerns the evidence of closure for the latest findings: does each one name a cause within the system, or a person?

There remains a grey zone that no one settles in writing: the threshold beyond which a series of minor deviations becomes a failure of the function that should have detected them. It depends on the inspector, the context and what the operator has previously been able to show. The only way to control it is to occupy it yourself: qualify the recurrence, name it in an internal finding and address its cause before the authority does so with its own scale.
For anyone aiming at a nominated-person role, or reviewing a hiring file, four questions are enough: date and scope of the last authority audit, number of open findings and their age, reporting line of compliance monitoring, existence of a deputy for each nominated post. Concrete cases are discussed on the FL330 forums; preparation for management and audit roles is on the training side.
This article describes comparative regulatory mechanisms and does not replace any official source. The text applicable in the state that issued the certificate, the operations specifications, the operator’s approved documentation and the decisions of the competent authority prevail over what is written here.
Questions fréquentes
Can a Level 2 finding become a Level 1 finding?
Yes, and it is the most common escalation path. The shift does not depend on the initial severity of the deviation but on how it is handled: a missed deadline, a rejected action plan, or a closure that does not address the cause are enough to reclassify the finding. The authority then reads it as a failure of the management system itself, which calls for immediate action and can directly affect the operations specifications.
Can a nominated person be replaced without waiting for the authority?
Not as a general rule: the authority accepted these people by name and must rule on the competence of the replacement. What can be handled internally is an interim arrangement, provided the change-management procedure has planned for it, that it is time-bound and notified. An interim arrangement that settles in without a file is one of the easiest grounds for a finding, since it shows up on an org chart.
Does a ramp inspection abroad have a consequence for the certificate?
The finding does not stay at the stopover: it is forwarded to the authority of the operator’s state and feeds into the oversight file. An isolated deviation carries little weight; it is the accumulation, and above all the repetition of the same deviation on different aircraft, that triggers a review of the internal system. The crew therefore has an interest in reporting it the same day, before the operator learns of it from its authority.
Does reporting a deviation yourself expose you to a sanction?
Both systems protect reporting, through different routes. The European occurrence-reporting regime prohibits using a report for punitive purposes against an unintentional deviation and protects the reporter’s identity. On the American side, a voluntary disclosure programme suspends enforcement action when the operator reports, stops the deviation immediately and gets a corrective plan accepted. In both cases, protection falls away if the deviation is deliberate or recurs after correction.
Can a certificate be withdrawn partially?
It is in fact the most common form of a severe action. Rather than stopping an operator, the authority acts on the operations specifications: it freezes a particular approval, removes an aircraft from the list, closes a base or restricts an area. Operations continue, but the commercial scope shrinks immediately, and restoring it requires a new demonstration, not a simple correction.
Can the compliance monitoring function be outsourced to a provider?
It can be carried out with outside support, particularly at smaller operators, but responsibility remains with the accountable manager and the arrangement must be described in the approved documentation. What the authority checks is independence from the activities being audited, direct access to the accountable manager, and the real capacity to cover all domains within the planned cycle. A contract that buys an annual audit with no follow-up on actions does not satisfy the requirement.

Leave a Reply
You must be logged in to post a comment.