Drone Regulation Index 2026 — methodology

The duties that attach to an ordinary, low-risk flight: registration, competency, standard altitude allowance and the everyday operating limits.

EFF-01 · EFF-02 · EFF-03 · EFF-04 · EFF-05 · EFF-06

Whether operations beyond the ordinary case — beyond visual line of sight, over people, at night, autonomous — have a published, reachable route rather than being simply unavailable.

ADV-01 · ADV-02 · ADV-06 · ADV-07 · ADV-08

How much process an operator carries: authorisation routes, validity periods, recognition of other markets' credentials and the weight of ongoing obligations.

ADM-01 · ADM-02 · ADM-05 · ADM-06 · CLR-02 · CLR-03

How the airspace itself is opened or closed: geographical zones, controlled-airspace access, and the conditions attached to flying near people, aerodromes and sensitive sites.

AIR-01 · AIR-02 · AIR-03 · AIR-04

Proportionate everyday duties, a reachable route to advanced operations, and administration and airspace access that a compliant operator can navigate.

A complete and workable framework, with either the advanced-operations route or the administrative path carrying more friction than the leading group.

The framework is in place, but several of the measured pathways are gated, narrow or hard to navigate for an operator seeking to comply.

The published framework resolves comparatively few of the measured pathways, or resolves them restrictively across most pillars.

Whether a recreational or light-commercial operator must register, and whether that duty is triggered proportionately (by weight, class or operation) rather than applying to every flight regardless of risk.

Direction:

Higher = registration exists but is proportionately triggered, or is not required at all.

Conditions:

Scope-limited conditions (weight/class/category) score 0.75; a permission gate scores 0.40; unclassifiable conditions score 0.50.

Not applicable:

Never N/A: every framework either imposes a registration duty or does not.

Evidence required:

registration.required must resolve on a binding basis or a structured threshold.

• A universal duty scores 0.25 rather than 0: registration is a proportionate safety measure, not a barrier.

How much competency proof an ordinary low-risk operator must obtain before flying: self-declared competency, training, or a formal examination.

Direction:

Higher = fewer stacked entry requirements before a first lawful flight.

Conditions:

Each component uses the shared obligation ordinal; the metric averages the three.

Not applicable:

Never N/A.

Evidence required:

At least two of the three components must resolve.

• A short online test and a multi-day course score identically: duration is not in the corpus.

The height an operator may use without any additional permission, in metres above ground.

Direction:

Higher = more usable vertical envelope by right.

Conditions:

A numeric value attached to a CONDITIONAL limit is used as-is: the condition is the limit itself.

Not applicable:

Never N/A.

Evidence required:

An explicit numeric value with a metre unit.

• Banded 60-150 m: differences inside the band are real but small, and the band is capped so an unusually permissive outlier cannot dominate the pillar.

Whether flying at night is available, and whether it needs a separate permission.

Direction:

Higher = night flight is available with fewer additional permissions.

Conditions:

Availability ordinal; a lighting condition is scope-limited, not a gate.

Not applicable:

Never N/A.

Evidence required:

permitted or authorisation_required must resolve.

• Where only the lighting rule resolves, the cell falls back to the authorisation dimension.

Whether operations over or near people who are not part of the flight are possible, and on what terms.

Direction:

Higher = a usable route exists for operating around people.

Conditions:

Aircraft-class or separation conditions are scope-limited; a permit is a gate.

Not applicable:

Never N/A.

Evidence required:

One of the three dimensions must resolve on a binding basis.

• An express prohibition scores 0 only where it is unqualified.

Whether commercial work is treated under the same risk-based rules as any other flight, or is pushed into a separate permit regime.

Direction:

Higher = purpose of flight does not by itself add a regulatory layer.

Conditions:

Separate treatment triggered only by operation category is scope-limited; a blanket commercial permit is a gate.

Not applicable:

Never N/A.

Evidence required:

separately_regulated or additional_permit must resolve.

• Risk-based frameworks that regulate by operation rather than by purpose score at the top of this metric by design.

How far beyond bare legality the framework goes for beyond-visual-line-of-sight flight: whether there is a named authorisation route, a risk-assessment method, a predefined or standard scenario, and repeat or generic permissions.

Direction:

Higher = a more complete, more repeatable and more predictable route to BVLOS.

Conditions:

The ladder consumes states directly. An unqualified FALSE on legality is a prohibition; a FALSE that coexists with a resolved authorisation route is read as a category-scoped prohibition (open-category style), not a national ban.

Not applicable:

Never N/A: every framework either offers a BVLOS route or does not.

Evidence required:

At least one of legality or authorisation type must resolve on a binding basis.

• Rung 5 requires an explicit repeat/generic or predefined route; frameworks that grant these in practice without expressing them in law score at rung 4.

Whether there is a named authority, a defined application type and a stated validity period for authorising operations beyond the everyday category.

Direction:

Higher = the route is fully specified in law rather than left to practice.

Conditions:

Presence of the element is what counts; conditions do not reduce the score.

Not applicable:

Never N/A.

Evidence required:

At least two of the four elements must resolve.

• Measures whether the route is specified, not how quickly the authority actually decides.

Whether airworthiness or type certification is demanded only for the operations that warrant it, rather than as a general precondition.

Direction:

Higher = certification is targeted at higher-risk operations.

Conditions:

An operation-category trigger is scope-limited (proportionate); a blanket duty is not.

Not applicable:

Never N/A.

Evidence required:

aircraft_certification.required must resolve.

• Cannot distinguish a light class-marking regime from full type certification.

Whether the higher pilot licence or certificate is tied to the operation being flown rather than required across the board.

Direction:

Higher = certification demanded in proportion to the operation.

Conditions:

As EFF-02, using the shared obligation ordinal.

Not applicable:

Never N/A.

Evidence required:

pilot_certification.required must resolve.

• Does not capture the cost or duration of obtaining the certificate.

Whether advanced operations such as inspection or delivery can lawfully take place over built-up areas.

Direction:

Higher = urban operations have a route.

Conditions:

Availability ordinal.

Not applicable:

Never N/A.

Evidence required:

The dimension must resolve.

• Presence of a route says nothing about how often it is granted.

How many separate registrations an operator must maintain — operator, aircraft, or both.

Direction:

Higher = fewer parallel registers to keep current.

Conditions:

A register triggered only above a threshold counts as half a layer.

Not applicable:

Where operator registration resolves FALSE and aircraft registration is therefore NOT_APPLICABLE, the cell still scores: the absence of the layer is the finding.

Evidence required:

At least one of the two must resolve.

• Does not capture the fee or the renewal interval; ADM-02 covers renewal separately.

Whether registrations and competency proofs expire and must be periodically renewed.

Direction:

Higher = fewer recurring renewal obligations.

Conditions:

Shared obligation ordinal across the three components.

Not applicable:

Never N/A.

Evidence required:

At least one component must resolve.

• A five-year validity and a one-year validity score identically where the period is not machine-readable.

Whether third-party insurance is mandatory, and whether proof must be filed.

Direction:

Higher = insurance is targeted rather than universal with filing duties.

Conditions:

A commercial or weight trigger is scope-limited.

Not applicable:

MISSING_DATA in four markets where the topic has no published fact; those cells are unresolved, never zero.

Evidence required:

insurance.required must resolve.

• Mandatory insurance is a consumer-protection measure; the metric is weighted as a small administrative cost, not as a regulatory failing.

Whether registration and authorisation can be completed online rather than on paper or in person.

Direction:

Higher = more of the process is available digitally.

Conditions:

Availability ordinal per component, averaged.

Not applicable:

Never N/A.

Evidence required:

At least two components must resolve.

• Measures existence of a digital channel, not its usability.

Whether the law names the authority that decides, the form the application takes, and where incidents are reported.

Direction:

Higher = an operator can identify who to deal with without external advice.

Conditions:

Presence counts; conditions do not reduce it.

Not applicable:

Never N/A.

Evidence required:

At least two of the three must resolve.

• Near-constant across the cohort: nearly every framework names its authority. Retained for completeness but it carries almost no discriminating power.

Whether exemptions and carve-outs are stated explicitly enough that an operator can tell whether one applies to them.

Direction:

Higher = exceptions are expressed rather than left implicit.

Conditions:

A resolved exception dimension counts whether TRUE, FALSE or CONDITIONAL: a clear 'no exceptions' is also clarity.

Not applicable:

Never N/A.

Evidence required:

At least two of the three must resolve.

• Rewards explicitness, which correlates with codification style as well as with clarity.

Whether the state publishes a defined geographical zone framework so an operator can know where restrictions apply.

Direction:

Higher = restrictions are defined and published rather than ad hoc.

Conditions:

A framework limited to certain zone types still counts as a framework.

Not applicable:

Never N/A.

Evidence required:

Either dimension resolved on a binding basis.

• Near-constant: 30/31 markets publish a zone framework, so this metric confirms a baseline rather than separating markets.

Whether there is a defined way to obtain access to controlled airspace and airport surroundings, and whether it can be done digitally.

Direction:

Higher = access is obtainable through a defined, ideally digital, process.

Conditions:

A permission requirement is not a penalty: controlled airspace is controlled everywhere. What scores is whether the route exists and how it is served.

Not applicable:

Never N/A.

Evidence required:

The controlled-airspace or airport dimension must resolve.

• Cannot measure approval rates or turnaround time.

Whether short-notice airspace restrictions are published through an official service an operator can check before flight.

Direction:

Higher = temporary restrictions are discoverable before flight.

Conditions:

Availability ordinal on the official system dimension.

Not applicable:

Never N/A.

Evidence required:

The official-system dimension must resolve.

• Scores the existence of an official service, not whether Drone Assistant can ingest it: a market is never penalised for our ingestion limits (I3 §21).

Whether an official map or service shows an operator where they may fly.

Direction:

Higher = a pre-flight map service exists.

Conditions:

Availability ordinal.

Not applicable:

Never N/A.

Evidence required:

Either dimension resolved.

• Deliberately independent of geometry ingestion: an official map the platform cannot parse still scores in full.

Whether an operational risk assessment method is available for higher-risk operations.

Consumed as rung 3 of ADV-01. Scoring it again would count one regulatory feature twice inside the same pillar (I3 §24).

Whether a predefined risk assessment or standard scenario removes the need for a bespoke case.

Consumed as rung 4 of ADV-01 (I3 §24).

Whether an operator can hold a standing permission instead of applying per flight.

Consumed as rung 5 of ADV-01 (I3 §24).

Whether a traffic-management or U-space framework is in force.

Resolves in 12/31 markets, far below the 80% acceptance gate; scoring it would penalise markets for silence rather than for regulation.

How many distinct permissions a typical advanced operation must assemble.

Three of its four dimensions are already scored inside Airspace Accessibility (AIR-01, AIR-02) and ADV-02; the fourth is constant across all 31 markets. Reported, not scored.

Whether registration is completed online.

Same feature as ADM-06's first component; reported here, scored once there.

Whether authorisations and flight plans are filed digitally.

Overlaps ADM-06 and AIR-02; reported to avoid triple counting.

Duplicate view of ADV-09 inside the digital panel.

Below the coverage gate at 12/31, and identical to ADV-09.

Whether remote identification is required and supported by infrastructure.

Has no defensible direction. Requiring remote ID is neither more nor less enabling; treating it either way would embed a policy preference in the Index (I3 §1).

Whether the state publishes its drone regulatory data in an open or machine-consumable form.

Resolves in 7/31 markets — the weakest metric in the register.