MULTI-AIRCRAFT SUPERVISION
How one person can maintain effective oversight as the number of aircraft and simultaneous missions grows.
Manage
RESEARCHFleet Management
Keep increasingly automated fleets manageable for humans.
Flyvercity researches the supervision, function allocation and operating architecture required when one aircraft becomes many.
The goal is bounded, observable automation—not autonomy detached from operational responsibility.
FUTURE OPERATING MODEL
OPERATORObjective + oversight
CONSTRAINTSFleet
intelligenceEXCEPTIONS
intelligenceEXCEPTIONS
UAS 01UAS 02UAS 03UAS 04
The market transition
The next constraint is not aircraft count. It is supervisory complexity.
Adding aircraft to today’s mission interface does not create a fleet operating system. Multiple simultaneous operations change workload, function allocation, failure management and the evidence expected of critical automation.
Flyvercity works on the layer between operating concept and software architecture: who or what performs each function, within which boundaries, and how the result can be observed and verified.
AIRCRAFT-CENTREDFLEET-CENTREDAircraft · routes · missionsObjectives · allocation · coordination · supervision
HIVE
Defining the requirements for future fleet management.
RESEARCH
HIVE investigates multi-aircraft supervision, function allocation, fleet coordination, separation, U-space interfaces, responsibility and degraded operations.
01Operating concept
02Supervision model
03Function allocation
04System requirements
05Evaluation
FUNCTION ALLOCATION
Which operational functions belong with people, deterministic automation or increasingly capable software.
FLEET COORDINATION
How missions, aircraft, constraints and priorities can be managed as one operational system.
HUMAN–AUTOMATION INTERACTION
How the system keeps intent, state, limits and exceptions understandable to the operator.
TACTICAL COORDINATION
How near-term interactions can be coordinated within the wider fleet-management architecture.
AI-ENABLED MANAGEMENT
How operators can express objectives while software handles more planning complexity within explicit constraints.
Market + compliance context
There is no single shortcut from one pilot to an autonomous fleet.
Current UAS authorisation remains tied to a defined operation, responsibilities, risks and evidence. Multiple simultaneous operations add a system-level question: how supervision, delegated functions, contingencies and service dependencies remain safe and understandable.
HIVE is research into that transition—not a generic approval route. As higher automation and AI enter operational systems, EASA’s human-centric AI work provides an important direction for learning assurance, explainability and human–AI teaming.
EASA / SORA 2.5Easy Access Rules for UAS
The June 2026 consolidated rules include the current SORA framework, terminology and operation-specific authorisation context.
OPEN SOURCE ↗SESAR JU / HIVEHigher fleet automation through HIVEEuropean research into multiple simultaneous operations, layered delegation and verifiable safety and performance requirements.
OPEN SOURCE ↗EASA / AI ROADMAPHuman-centric AI in aviationEASA’s living roadmap frames AI around safety, security, human factors and a risk-based path to higher automation.
OPEN SOURCE ↗EASA / AI ISSUE 2AI Concept Paper — Issue 2Guidance concepts for learning assurance, explainability, ethics and human–AI teaming for Level 1 and Level 2 applications.
OPEN SOURCE ↗These sources describe the current framework and research direction; they do not establish a universal approval path for fleet autonomy. Flyvercity develops operating concepts, architectures and verifiable requirements to support future cases, while approval remains specific to the operator, operation, system and competent authority.
Market + compliance trajectory
From aircraft-centred control to objective-led fleet supervision.
This is a capability and assurance roadmap, not a prediction of regulatory dates. Each step increases the functions delegated to software—and therefore the need for explicit boundaries, observability and evidence.
EVOLUTION
TODAYSINGLE OPERATIONS
EMERGINGMULTIPLE SIMULTANEOUS OPERATIONS
U-SPACE INTEGRATEDDELEGATED FUNCTIONS
LONGER TERMOBJECTIVE-LED FLEETS
MARKET
01
Most operational tools remain aircraft- and mission-centred, with one remote pilot responsible for one active operation.
02
One operator or team supervises several aircraft and concurrent missions through shared fleet workflows.
03
Fleet systems take on more planning, conformance, coordination and exception-handling functions within controlled boundaries.
04
Operators express objectives while software allocates aircraft and manages more of the underlying execution complexity.
COMPLIANCE
01
The authorisation and SORA evidence are built around a specific operation, explicit responsibilities and defined contingency procedures.
02
Responsibility, workload, intervention, contingency management and system performance need explicit, verifiable substantiation.
03
Service interfaces, authority boundaries, performance assumptions and human–automation interaction become part of the assurance case.
04
Higher automation and AI introduce learning-assurance, explainability, human-factors and trustworthiness expectations that are still evolving.
FLYVERCITY
01
Baseline the operating concept, human tasks, information needs, constraints and degraded states before increasing supervisory scale.
02
HIVE research: define supervision models, function allocation, safety requirements and evaluation scenarios for MSO.
03
Design bounded automation architectures, U-space interfaces, observability and repeatable tests for operationally critical functions.
04
Research constrained, observable fleet intelligence in which human intent, limits, reasoning and managed exceptions remain explicit.
AI-ENABLED FLEET MANAGEMENT
Make increasingly capable fleet systems easier to operate.
Flyvercity is exploring how AI can help operators express operational objectives while software manages more of the underlying planning and operational complexity within explicit constraints.
The objective is not unrestricted AI control. It is to make powerful autonomous systems easier for people to supervise, understand and use.
01Operational objective
02Fleet intelligence
03Constraint checks
04Verified execution
Human intentExplicit constraintsObservable reasoningManaged exceptions
Research principles
Automation that remains operationally accountable.
HUMAN-CENTRED
Operator roles, workload, awareness and intervention remain part of the system design.
CONSTRAINT-DRIVEN
Objectives are pursued inside explicit operational and safety boundaries.
DEGRADED-STATE AWARE
Responsibility and behaviour remain clear when services, aircraft or information degrade.
VERIFIABLE
Future capabilities need testable requirements and observable operational evidence.
Programme context
Research connected to real UAS programmes.
Flyvercity’s work combines systems engineering, software development, operational experimentation and European research collaboration.
HIVESESARIoT-NGINHORIZON 2020SAFIR-ReadySESAR
Start with the operation
Working on the future of fleet-scale UAS operations?
We are interested in operating concepts, fleet-supervision architectures and research collaborations grounded in real operational constraints.
Discuss research collaboration ↗