Manav.id
Vertical · 5 min read

The dispatch release and the MEL deferral: two signatures that decide whether a flight is legal

The dispatch release and the MEL deferral: two signatures that decide whether a flight is legal

Two decisions determine whether a flight may legally depart. A dispatcher and a captain jointly exercise operational control, and a maintenance deferral under the minimum equipment list makes an aircraft with a broken component legal to carry passengers. Both used to be signatures on paper. Both are now workflow states.

Which two signatures make a flight legal?

The dispatch release, exercised jointly by the captain and the dispatcher under domestic operating rules, and the MEL deferral that makes an otherwise unairworthy aircraft legal to fly. Both are recorded as system states in operational software, and neither is bound to the specific configuration it certifies.

Key takeaways
  • Joint operational control is a shared legal responsibility. Shared responsibilities are precisely the ones that produce unresolvable disputes when the record names accounts rather than certificated individuals.
  • An MEL deferral is airworthiness relief. It is frequently applied outside base maintenance by delegated staff under time pressure at the gate.
  • A dual-principal receipt over the rendered release packet and deferral basis costs seconds and removes the ambiguity.

The uncomfortable structure of shared authority

MEL deferralitem inoperativemakes it legalDispatch releasecaptain + dispatcherjoint controlFlight departsunder bothlegallyRecord: two status fieldsin the ops systemnames accounts
Joint operational control is a legal construct. The evidence for it is two application states.

Joint operational control is one of the more elegant ideas in aviation regulation. Neither the dispatcher nor the captain can unilaterally launch a flight; both must concur, and both carry responsibility. The design assumes that two people, with different information and different incentives, will catch what one would miss.

It also has a failure mode that regulation did not anticipate, because digitisation created it. When responsibility is shared and the record is weak, the shared responsibility becomes a shared inability to establish what either party actually did.

Paper release packets were signed. Electronic ones are acknowledged. Those are different acts wearing the same name.

What the release actually is now

The dispatcher builds the release: route, fuel, weather, alternates, weight and balance, notices. It is transmitted to the aircraft. The captain reviews it on an electronic flight bag and acknowledges.

The acknowledgement is a button on a company-managed tablet, frequently logged into a company account, sometimes shared across a crew base. The record shows an acknowledgement occurred. It does not show that the captain read the fuel figure, the alternate, or the notice about the inoperative component.

The MEL deferral is the sharper problem

A minimum equipment list deferral says: this component is inoperative, and under these conditions, with these operational and maintenance procedures applied, the aircraft may be dispatched anyway.

That is airworthiness relief granted by a human under time pressure at a gate, often by maintenance control remotely, often communicated by phone, and recorded as an entry with a category and an expiry.

Element of the deferralRecorded todayEvidentiary strength
Which MEL itemYes, by referenceGood
Category and expiryYesGood
That the (O) and (M) procedures were appliedAs a checkbox or a separate maintenance entryWeak — not bound to the deferral
Who authorised the deferralUser id in the maintenance systemWeak — session-attributed, often a shared maintenance control account
That the captain was informed of the deferral before acceptanceImplied by the release packetNot evidenced

Why the fifth row is the one that matters

Every serious dispatch-related investigation eventually asks whether the captain knew. The answer today is derived: the deferral was on the release, the release was acknowledged, therefore the captain knew.

That inference chain is only as strong as its weakest link, and the weakest link is an acknowledgement button on a shared tablet. A captain who says they never saw the deferral cannot be contradicted, and a captain who did see it cannot be exonerated.

Shared authority with weak attribution is not two people checking each other. It is two people unable to establish what either did.

The Operational Control Receipt

One canonical statement, two certificated signatures, bound to the rendered content.

{
  "type": "manav-stmt/1",
  "action": "dispatch_release",
  "render": [
    "Flight: [number]  Aircraft: [registration]  Date: [date]",
    "Route: [origin]–[destination]  Alternates: [list]",
    "Fuel: [release fuel]  Payload: [figure]",
    "MEL items open: [item] cat [A-D] expires [date]",
    "Applied procedures: (O) [ref]  (M) [ref]"
  ],
  "signers": ["dispatcher [cert]", "pilot in command [cert]"]
}

The deferral appears inside the rendered content, so the captain's signature covers it explicitly. The inference chain collapses into a single verifiable fact.

The turnaround objection, answered with a number

Flight operations will object on tempo, and they are right to. A control that adds thirty seconds to a turnaround at a hub does real damage across a day of operations.

So measure before proposing. Instrument the gesture on the actual EFB hardware in use, at a gate, with a line captain, under normal pre-departure workload. Report median and p95. A control with a median of roughly a second and a p95 under four is a different conversation from one that occasionally takes fifteen because the tablet is negotiating a network it cannot reach — which is why offline verification matters here specifically.

What stays the same

Everything procedural. The dispatcher builds the release the same way. Maintenance control grants the deferral under the same authority. The captain reviews the same packet. No new approval step is introduced, and nothing blocks a departure that is currently permitted.

The only change is that the two confirmations already being made produce records that name certificated individuals and cover the content those individuals confirmed.

What the record should bind

The elements that decide legality
ElementWhy it must be inside the signature
Aircraft registration and configurationA deferral applies to one aircraft in one state
MEL item, category and expiryA deferral has a clock; an expired one is not a deferral
Operational limitations imposedThe conditions under which the deferral holds
Release parametersFuel, alternates, weather window
Both named individualsJoint control means two people, not two accounts

Objections and honest limits

“The ops system already records all of this.” It records it. The question is whether the signature is bound to that configuration, so that a change after release invalidates it rather than silently proceeding.

“Dispatchers release dozens of flights a shift.” Which is exactly why the signature must be cheap — one gesture on the release — and why it should render the deltas from standard rather than the whole release packet.

Binding the release

  1. Render the deltas, not the packet. What is inoperative, what limitation applies, what changed from standard.
  2. Bind both signatures to the same configuration digest. So a change between them invalidates the first.
  3. Carry the MEL expiry inside the signed statement. An expired deferral should fail verification, not require a human to notice.
  4. Recompute at departure. A configuration change after release refuses rather than flies.

Terms used here

Dispatch release
The document authorising a flight, under domestic operating rules jointly the responsibility of the captain and the dispatcher.
MEL
Minimum Equipment List: the approved basis on which an aircraft with an inoperative item may be operated, subject to limitations and a repair interval.
Operational control
The authority over initiating, conducting and terminating a flight — a legal concept with named holders.

Frequently asked questions

Does this add an approval step? No. It changes what the existing acknowledgements produce. Nobody new is inserted into the decision.

What if the EFB has no connectivity at the gate? Verification is offline against a published key held on the device, which is a deliberate design choice for exactly this case.

Who signs when maintenance control grants the deferral remotely? The authorised individual in maintenance control signs the deferral, and the captain separately signs the release that carries it. Two distinct acts, correctly attributed.

Is this required under Part 121? No. Part 121 establishes operational control and MEL authority. It does not prescribe the evidentiary form of the resulting records.

Why do two signatures matter here? Because operational control is exercised jointly. Two application states naming two accounts is not the same as two people each attesting to the same configuration.

What should the deferral signature cover? The aircraft, the item, the category, the expiry and the operational limitations — so an expired or changed deferral fails verification.

Does this slow dispatch? It should not. One gesture on the release, rendering the deltas from standard rather than the full packet.

Where this fits in Manav

Manav binds an airworthiness or operational signature to the exact record it certifies, on a credential under one individual's sole control, and produces a receipt an auditor or a regulator can verify years later without access to the operator's systems.

See signature binding →

Sources and further reading