Guide 01 / Mission and input control

Set the boundaries

An optimisation problem begins with a boundary. State what the aircraft, operation and passenger promise permit before asking software what could fit.

For fleet, cabin-product and planning teamsRead next: passenger comfort

Write an operating mission

Start with the service the cabin must deliver. Describe sector duration, rotation pattern, turnaround pressure, cabin service, expected baggage and the passenger groups you intend to accommodate. A short sector with frequent boarding may place different demands on stowage and circulation than a flight on which passengers expect prolonged rest. Neither mission automatically determines the right seat count.

Separate a planning scenario from a demand forecast. If the analysis assumes strong premium demand, label that assumption and name its source. Do not present a commercially preferred cabin split as a physical necessity. Record who can change the mission and which decisions would trigger a new comparison.

Aircraft names alone are insufficient input identifiers. Airbus A320 and Boeing 737 families contain different variants and configurations. Even aircraft within a nominally common fleet can have different installed equipment, modification histories or approved arrangements. Use the applicable aircraft and configuration records rather than dimensions copied from a general fleet description.

Define geometry and interfaces

Establish a controlled coordinate system and consistent units before importing geometry. Identify the cabin envelope, floor and seat-track references, doors, exits, monuments, lavatories, galleys and fixed equipment. A layout drawing that combines different reference origins can look convincing while placing a component in the wrong position.

A seat is not just a rectangular footprint. Capture relevant installed envelopes, including recline, armrest movement, leg support and access to controls. Include interfaces for power, in-flight entertainment and other systems where they affect the comparison. State whether the software checks these relationships directly or merely displays geometry that an engineer must inspect.

Mass and balance also need explicit treatment. More seats, relocated monuments or changed equipment can affect mass distribution and operating assumptions. A visually feasible arrangement does not establish acceptable aircraft loading. Identify the responsible analysis and avoid implying that a geometric placement engine covers it automatically.

Close-up of an overhead cabin fitting with reading light and ventilation controls
Overhead fitting: local interfaces belong in the input register when they affect reach, service or installation.

Separate rules from preferences

A hard constraint is a condition that a candidate must satisfy within the stated model. A preference helps rank candidates that remain eligible. Mixing the two can let a weighted score compensate for something that should never be traded away. For example, a commercial benefit must not cancel an applicable safety requirement.

Translate requirements into traceable checks rather than vague labels such as “compliant cabin”. EASA CS-25 and FAA 14 CFR Part 25 provide transport-category airworthiness frameworks, but applicability depends on the aircraft certification basis and proposed change. The planning team should not treat an isolated limit copied from a standard as a complete compliance assessment.

Assumption note: a software default is not a requirement

Default aisle, seat or service-space values may be starting assumptions. Record their origin and obtain the relevant technical review before treating them as controlling limits. A supplier’s label does not replace an aircraft-specific requirement.

Assign each constraint an owner and a review state. Cabin-product teams may own an experience target, while engineering controls an installation boundary and operations owns a turnaround assumption. Disagreement is useful information. Keep unresolved values visible instead of averaging them into an apparently agreed input.

Make uncertainty portable

Build an assumptions register that can travel with every exported layout. Include the input name, value or range, units, source, configuration applicability, owner and review date. Mark measured values, supplied values and provisional estimates distinctly. Confidence in one input must not be mistaken for confidence in the entire model.

For uncertain quantities, use plausible ranges and test whether a preferred option changes. If a small adjustment to baggage assumptions reverses the ranking, the conclusion is conditional. That is a planning result worth preserving, not an inconvenience to hide in an appendix.

Before proceeding, ask another team member to reconstruct the baseline from the register. If they need undocumented explanations, improve the record. Reproducibility is a practical quality check: it reveals ambiguous units, missing configuration details and preferences disguised as facts.

Register fieldRecordReview question
MissionSector and service assumptionsWhose operation is represented?
GeometryControlled source and revisionDoes it match the aircraft?
ConstraintRule, owner and applicabilityIs it hard or preferred?
UncertaintyRange and evidence gapCould this change the decision?

Decision gate: every consequential input has a source or an explicit uncertainty label. Unknown does not mean unrestricted.