91.147 air tour operators
have an SMS deadline too.

See the highlighted block above the contents list. The rest of this article walks through exactly who the 2024 rule pulled in, the two dates that bind you, the hazards a tour operation actually runs, why reporting and a seasonal workforce need planning, and a pragmatic path to full implementation by 2027.

For years, a formal safety management system was an airline obligation. The FAA SMS final rule changed that. It was published on 2024-04-26 and took effect on 2024-05-28, revising 14 CFR Part 5 and extending it well beyond Part 121. The rule now reaches all Part 135 certificate holders, certain Part 21 certificate holders, and, the part that catches many operators by surprise, commercial air tour operators conducting operations under a 14 CFR 91.147 letter of authorization.

If your business is sightseeing flights sold to the public, the helicopter run over a canyon, the floatplane circuit around an island, the fixed-wing loop over a coastline, and you hold a 91.147 LOA to do it, you are a covered operator. The wider timeline and the reasoning behind the extension are laid out in our Part 5 final rule explainer; this article is about what it means for a tour operation specifically.

The most important word in the rule for a tour operator is “regardless.” The obligation applies regardless of operator size, and the FAA states it explicitly for single-aircraft and single-pilot certificate holders. There is no headcount below which the rule stops. A one-ship seasonal operator and a large multi-base tour company are both inside it. What differs is not whether you need an SMS, but how large that SMS has to be.

Two milestones matter, and they are the same ones that apply to Part 135 operators under the rule:

• 2024-11-28, implementation plan. Covered operators were to submit an implementation plan to the FAA, the roadmap that describes how and when the SMS will be stood up.
• 2027-05-28, full implementation. The SMS must be fully implemented and a declaration of compliance submitted to the FAA no later than this date.

The declaration of compliance is the formal, dated statement that your SMS meets the rule. It is not paperwork you generate at the end from nothing; it is the moment you certify that a working system already exists and can be evidenced. Our declaration of compliance checklist covers what has to be true before you sign it. For the exact submission mechanics that apply to your certificate and LOA, check the current FAA guidance rather than relying on a date alone.

Read the two dates the right way round. Full implementation is the milestone that consumes real calendar time, because a functioning SMS needs reports flowing, hazards logged, risk decisions recorded, and an assurance cycle that has actually turned over at least once. That evidence cannot be assembled in the final weeks. An operator who treats 2027-05-28 as the start of the work rather than the proof of it has misread the rule.

Safety Risk Management is the component where an air tour SMS earns its keep, because tour flying carries a distinctive hazard profile. A generic template written for scheduled transport will miss most of it. The tour-specific hazards worth naming first:

• Terrain and obstacle proximity. The product is the view, which means low-altitude flight close to canyon walls, ridgelines, coastlines and built obstacles. Margins that a cruise profile never touches are the daily working environment.
• Weather windows. Scenic areas generate fast local weather: valley fog, ridge turbulence, sea breezes, afternoon build-ups. The pressure to fly a booked, paid tour into a marginal window is a real and recurring hazard, not a rare one.
• High cycle counts. A good-weather day is many short flights back to back. High daily cycles compress turnarounds, stack fatigue across a long season, and raise the exposure of every ground and airborne task through sheer repetition.
• Shared scenic corridors. Popular routes concentrate multiple operators into the same airspace and the same photo lines, raising traffic and separation risk in exactly the places passengers most want to be.
• Passenger and ground handling. Non-crew passengers loading and unloading near running aircraft, often in rotor-wash or on unimproved sites, is a hazard the flight crew do not fully control from the cockpit.

The rule does not want a list of these on a shelf; it wants them identified, assessed, controlled, and monitored. A pre-flight flight risk assessment tool is the natural first control for the weather-and-terrain decision, turning “it looks flyable” into a scored, recorded call. eAviora ships a pre-built aviation risk model, 101 bow-tie models, 804 named barriers and 210 scenarios, so a small operator does not start the hazard register from a blank page: the terrain, weather and ground-handling scenarios are already modelled and ready to scope to your operation.

Hazard identification only works if the people who see the hazards will say so. The 2024 revisions make this concrete by requiring a confidential employee reporting system: a channel that protects the reporter's identity while feeding the risk process. For a tour operation, the reporters are not only pilots. Ground crew, loaders, dispatchers and reservations staff see near-misses on the ramp, the schedule pressure that pushes a marginal go, the passenger who ignored the safety brief, long before any of it reaches a flight deck.

Confidentiality is what makes the channel real. If a report can be traced back to the loader who filed it, the loader stops filing. A closed just culture loop is the answer: confidential intake, an investigation, a de-identified bulletin so the whole workforce learns, and feedback to the reporter that something happened, all without ever exposing who spoke. eAviora runs exactly that loop, with reporter identity protected by the platform, not by a promise.

Then there is the seasonal problem, which is sharper for tours than for most operators. A workforce that turns over between seasons means the safety culture, the reporting habit and the hazard knowledge walk out the door every autumn and have to be rebuilt every spring. An SMS that lives only in one veteran chief pilot's head does not survive that churn. The system has to hold the knowledge: the hazard register, the reporting channel, the training records and the lessons learned need to persist between seasons and onboard each new cohort quickly. This is as true for a fixed-wing operator as for the many helicopter tour operators in the same bracket, where seasonal crewing is the norm.

The trap is to over-build. A tour operator does not need the SMS of a legacy carrier, and trying to copy one produces a binder nobody uses. The rule asks for a system scaled to the size and complexity of the operation, so build the four components honestly and no larger. A workable order:

• Safety Policy first. Name the accountable executive, state the safety objectives, and set the rules of the system in a few pages people will actually read.
• Open the reporting channel early. A confidential channel that is live for a full season before 2027 gives you real reports to point at, not an empty inbox.
• Stand up the hazard register. Start from the tour-specific hazards above, score them, and record the controls you already run so the register reflects reality on day one.
• Turn the assurance cycle at least once. An internal audit, a management review, a corrective action driven to closure. One completed loop is worth more than a thick manual.

Tooling is where a small team stops drowning. When occurrences, hazards, investigations, corrective actions, audits, findings, training and safety performance indicators live in one operational graph, the same records that run the operation become the evidence for the declaration, with nothing re-keyed into a separate compliance binder. That is the model behind the eAviora SMS module: a pre-built risk model and a 610-indicator SPI library so you are not authoring metrics from scratch, a confidential just-culture loop, and enforced closure gates that will not let a record close over open risk, so a degraded barrier has to have a corrective action proven effective before anything reads as closed. AI drafting is available on the workflow, metered in plain credits with hard caps and human-gated on every proposal, and your operator data never trains any model.