Flying drones under Part 107

Airspace decides whether you may launch at all. Class B, C, D and the surface area of Class E designated for an airport all require prior authorization before you operate there. Class G — uncontrolled — needs none, which is where most small-drone work happens.

The trap is thinking in altitude. You are almost certainly flying below 400 feet, and that number appears in every summary, so it feels like the controlling fact. It is not: a 200-foot flight inside a Class D surface area still needs authorization, and a 380-foot flight in Class G does not. Position decides, then altitude.

Class E is where this bites, because it has two shapes. Where Class E extends to the surface at an airport, you need authorization. Where it begins aloft — at 700 or 1,200 feet above ground level (AGL) — the airspace under it is Class G, so a 400-foot job beneath an E-aloft floor needs nothing. Same letter, opposite answer, decided by the floor.

Beyond class, the airspace can be closed outright: temporary flight restrictions over stadiums, wildfires and VIP movements, plus standing national-security restrictions. Those are published, they change daily, and checking them is part of every preflight rather than an occasional errand.

Rule: before any flight, name the airspace class and its floor at your exact launch point — if you cannot, you are not ready to fly, whatever your altitude.

Aviation weather is published as a program, not a page. The National Weather Service and the FAA issue observations, forecasts and in-flight advisories, and each answers a different question. Observations report what a station is measuring now, and forecasts state what is expected over a period, for a place. Advisories warn of specific hazards, and they interrupt rather than wait to be looked up.

The misconception to clear early is that a consumer weather app substitutes for this. It reports rain and warmth. Aviation products report what makes flight unsafe: visibility, ceiling, wind, and the hazards that ruin a day quietly, like icing conditions or a line of convection moving in.

Ceiling-and-visibility categories were built for manned flight, but they compress "how much can you see, and how far up" into one word you can act on. Visual line of sight is your legal and practical constraint, so visibility is the limit you are flying against. Borrow manned aviation's preflight briefing too: a deliberate check before the day, not a glance at the sky when you arrive.

Concretely: the AIM's own sample observation reads METAR KBNA 281250Z 33018KT 290V360 1/2SM R31/2700FT SN BLSN FG VV008 00/M03 A2991. That is Nashville, the 28th at 1250 universal time. Wind is 330 at 18 knots, swinging between 290 and 360; visibility is one half mile in moderate snow, blowing snow and fog; the ceiling is indefinite at 800 feet; temperature is zero and dew point minus three. In the AIM's categories that is low IFR (ceiling under 500 feet and/or visibility under a mile), where visual flight rules (VFR) conditions mean a ceiling above 3,000 and visibility above 5. For a pilot who must keep the aircraft in sight, the half-mile is the whole verdict.

Rule: match the product to the question — observation for now, forecast for later, advisory for dangerous — and never let a general-purpose app answer any of the three.

Manned aviation cares about weather across thousands of feet. A drone lives in the first few hundred, the layer where terrain and surface heating do the most damage.

Wind is not one number. It increases with height and changes direction as ground friction falls away, so a 10-knot reading at your feet can be considerably more at 300 feet. The aircraft then has to fight it home on a depleted battery. Buildings and treelines add mechanical turbulence: rotor and shear downwind of an obstacle, sometimes violent, always invisible.

Heating drives the rest. Uneven surface warming produces thermals and gusty low-level air, which is why a calm morning and a bright afternoon at one site are different flying days. Convection escalates that into the one hazard nobody flies near: a developing thunderstorm produces gusts, outflow and downdrafts well outside the visible cloud.

Density is the quiet one. Hot, high and humid air is thinner, so the rotors move less mass per revolution. The aircraft needs more power just to hover, and the margin you counted on is not there.

Concretely: the AIM's microburst figures live at drone altitude. Downdrafts can reach 6,000 feet per minute. Near the ground they turn into an outflow roughly two and a half miles across, with surface winds up to 45 knots: a 90-knot swing from headwind to tailwind for anything crossing it, within a few hundred feet of the ground. Its advice is written for airliners and applies harder to you: never regard any thunderstorm lightly, and do not fly under one even when you can see through to the other side.

Rule: read wind, heating and density as one question — how much thrust margin will I actually have at altitude — because each of the three reduces it in a different way.

Payload is a trade, and the currency is thrust margin. Extra weight raises the power you need just to stay level, which at once shortens endurance, flattens climb, and shrinks your margin to recover from a gust. The trade steepens near maximum weight: the last hundred grams cost more than the first.

Balance is separate from weight, and it is the one people skip. An off-center load makes a multirotor hold a constant correction, visible as a persistent lean or drift, which burns power and eats margin before anything has gone wrong. Fix it on the bench, not in the air.

Battery discipline is the other half. Voltage sags under load, so remaining percentage is not linear: the last quarter goes faster than the first, and fastest exactly when you are working hardest. Cold reduces usable capacity outright; heat ages the pack and raises its failure risk.

The classic trap is a tailwind out and a headwind back. The outbound leg feels quick and efficient; then you return into wind, at higher power, on the depleted end of the pack.

Concretely: the advisory circular warns that even with a published maximum gross takeoff weight, the aircraft may not safely take off with that load under all conditions. High elevations, high temperatures and high humidity (high density altitude) may require a reduction in weight before flight. So a survey rig loaded to the limit at sea level in spring is over the limit on a mountain site in August, with the same number on the scale. Its risk list names the pairing outright: configuration means gross weight and center of gravity.

Rule: plan the reserve for the return leg at the end of the pack in the day's conditions — never from how the outbound leg felt.

Aeronautical decision-making (ADM) is described as a systematic mental process for consistently determining the best course of action. That word is the point: consistency comes from running a procedure, not from feeling careful.

Crew resource management sits inside it, and it is broader than "talk to your team". It means making effective use of every available resource — people, hardware and information — which for a drone crew includes the visual observer, anyone manipulating the controls, and the bystanders who wander into your site. The named skills are situational awareness, allocating tasks deliberately, avoiding work overload in yourself and in others, and communicating clearly with the crew.

The FAA names the hazardous attitudes and, more usefully, what they produce: getting behind the aircraft or the situation, operating without adequate battery reserve, losing positional or situational awareness, flying outside the envelope, and skipping planning, preflight or checklists. Operational pressure is named as the contributor — the client standing there, the schedule, the one more shot.

Drone-specific stressors are called out too, and they are unglamorous: an inexperienced crew member, no standard crew training, dealing with the public and city officials, and keeping up with changing rules.

Rule: set your minimums before the day starts and write them down, because the moment you need them is the moment the pressure to move them is highest.

Even if you never key a microphone, the air-traffic structure tells you where manned aircraft will be and how they behave.

The system divides by facility. Centers handle en-route traffic between terminal areas, and towers control aircraft in the immediate airport environment. Flight service exists for briefings and pilot information rather than separation.

What matters most for low-altitude work is what happens where nobody is in charge. At airports without an operating control tower, pilots self-announce on a common frequency and coordinate by convention rather than instruction. Such an airport may have a part-time aeronautical station or none at all, and monitoring the designated common frequencies is a legitimate way to know what is inbound. Traffic near a small field is predictable in pattern but not commanded, so "no tower" means more attention, not less.

The misconception: that authorization to fly in controlled airspace is a conversation. It is a prior authorization, obtained before the flight rather than by radio at launch, and holding it does not make you a participant in the traffic pattern.

Concretely: a light aircraft inbound to the non-towered field a mile from your site announces itself on the Common Traffic Advisory Frequency at ten miles out. It announces again entering downwind, base and final, then leaving the runway; where the field has no station at all, the AIM's frequency is 122.9. Monitor that frequency and the pattern tells you it is live before you can see it. Meanwhile the center is working IFR traffic en route and flight service is giving briefings; neither is watching your field.

Rule: identify which facility owns the airspace over your site and whether the nearest field is towered — that single fact predicts how the manned traffic around you will behave.

Before each flight, the remote pilot in command must ensure the aircraft is in a condition for safe operation, and the inspection includes examining it for equipment damage and malfunction. The obligation is per flight, not per day and not per job.

Where the manufacturer publishes inspection procedures, usually in the owner's or maintenance manual, those are the basis. Where it does not, you develop your own. Owners may go further and build an inspection program, drawing on industry programs for the detail.

The check should be scalable to the aircraft, the program and the operation actually flown: a short flight over open ground and a full day over a work site are not the same inspection.

The failure mode this prevents is treating visible damage as cosmetic. Propeller nicks and imbalance cost thrust and add vibration, and vibration is how a small problem becomes a structural one mid-flight. The same reasoning applies to a battery that has been dropped, swollen or deeply discharged.

Concretely: the advisory circular lists what the FAA wants in that checklist even when the manufacturer supplies one, twenty-three items. It starts with a visual condition inspection, then covers the airframe including undercarriage, control surfaces and linkages, and the registration markings for display and legibility. It checks that a camera or other equipment is securely attached, and that the control link is established and the aircraft has acquired GPS from the manufacturer's minimum number of satellites. Last come battery levels for the aircraft and the control station. Where no procedure exists, it points at Part 43 appendix D as the skeleton to build one on.

Rule: write the checklist once, keep it with the aircraft, and run it before every flight — a check you perform from memory is a check you will shorten under pressure.

The regulation frames emergencies as a briefing item. Everyone directly participating must be informed of the operating conditions, the emergency procedures, the contingency procedures — explicitly including those for people or moving vehicles entering the operational area — each person's role, and the potential hazards.

Read that list again and note what it assumes: the decisions already exist. The briefing distributes them; it does not invent them. An operator improvising a plan while an aircraft misbehaves is doing two hard things at once.

Control-link integrity is checked before flight rather than trusted, and the link is where most drone emergencies begin. Before each flight the crew confirms all control links between the control station and the aircraft are working properly.

The FAA's own worked example is worth the space: you have an obvious landing site, then a medical helicopter needs that same area while you are low on power. Thinking it through beforehand, you plan an alternate landing site, and further sites where, if it comes to it, you would sacrifice the aircraft rather than impose risk on people on the ground or on manned aircraft.

That is the mental move most operators never make. The aircraft is the cheapest thing in the situation.

Rule: decide before launch where you will put it down, where you will put it down instead, and what you are willing to lose — then brief all three.

Planning starts with identifying the hazards, which the FAA names as the first step of any aviation operation; mitigation follows, then you judge whether the residual risk is acceptable.

Visual line of sight organizes everything else. The aircraft must stay within VLOS of you or of a visual observer who is co-located and in direct communication with you, and as remote pilot in command you keep see-and-avoid responsibility throughout. Brief glances at the controller to check battery or maintain awareness are anticipated; if you lose sight, regain it as soon as practicable.

The aircraft must also give way to manned aircraft, which makes a landing decision a yielding decision. And as remote pilot in command you are directly responsible for the operation, which settles who decides.

Where a rule genuinely cannot be met, several operating rules are waivable through a defined waiver process. The trap is treating waivable as optional.

Concretely: a rooftop inspection over a busy sidewalk. Hazard first: people passing beneath the flight line. Then the advisory circular's two-part rating: how likely a failure is to put the aircraft on them, and how severe if it does. Then mitigations, which it says must reduce the probability of the cause, the severity of the effect, or both. A cordon and a crossing point away from the line cut the likelihood, a lighter aircraft cuts the severity, and a visual observer beside you keeps the aircraft in sight while you watch the sidewalk. Its list of what to weigh (workload, configuration, environment, the aircraft's limits, the consequence of failure) is the survey, and it suggests the crew brainstorm it together.

Rule: survey the site by standing where you will stand and listing what you cannot see — VLOS is the constraint the rest of the plan is shaped around.