See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/310316914

An Evolutionary Outlook of Air Traffic Flow Management Techniques

Article  in  Progress in Aerospace Sciences · January 2017

DOI: 10.1016/j.paerosci.2016.10.001

CITATIONS READS

20 2,693

5 authors, including:

Trevor Kistan Alessandro Gardi

RMIT University RMIT University
34 PUBLICATIONS   337 CITATIONS    139 PUBLICATIONS   1,540 CITATIONS   

SEE PROFILE SEE PROFILE

Roberto Sabatini Subramanian Ramasamy

RMIT University RMIT University
397 PUBLICATIONS   3,313 CITATIONS    122 PUBLICATIONS   1,606 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Next Generation Air Traffic Management and Avionics Systems View project

Trusted Autonomy in Future Aviation and Aerospace Systems View project

All content following this page was uploaded by Roberto Sabatini on 22 September 2017.

The user has requested enhancement of the downloaded file.

This is the author uncorrected pre-publication version. This paper does not include the changes arising from the revision, formatting and
publishing process. The final paper that should be used (available at http://dx.doi.org/10.1016/j.paerosci.2016.10.001 ) is:
T. Kistan, A. Gardi, R. Sabatini, S. Ramasamy and E. Batuwangala, “An Evolutionary Outlook of Air Traffic Flow Management Techniques”,
Progress in Aerospace Sciences, vol. 88, pp. 15-42, 2017. DOI: 10.1016/j.paerosci.2016.10.001

An Evolutionary Outlook of Air Traffic Flow Management Techniques

Trevor Kistan1,2, Alessandro Gardi1, Roberto Sabatini1,*, Subramanian Ramasamy1,, Eranga Batuwangala1
1
RMIT University, Melbourne, Victoria 3001, Australia; 2 Thales Australia Air Traffic Management, Melbourne, Victoria 3001, Australia

ARTICLE INFO ABSTRACT

In recent years Air Traffic Flow Management (ATFM) has become pertinent even in regions without sustained
Article history:
overload conditions caused by dense traffic operations. Increasing traffic volumes in the face of constrained
Received resources has created peak congestion at specific locations and times in many areas of the world. Increased
environmental awareness and economic drivers have combined to create a resurgent interest in ATFM as
Received in revised form
evidenced by a spate of recent ATFM conferences and workshops mediated by official bodies such as ICAO,
Accepted IATA, CANSO the FAA and Eurocontrol. Significant ATFM acquisitions in the last 5 years include South
Africa, Australia and India. Singapore, Thailand and Korea are all expected to procure ATFM systems within a
year while China is expected to develop a bespoke system. Asia-Pacific nations are particularly pro-active
given the traffic growth projections for the region (by 2050 half of all air traffic will be to, from or within the
Keywords: Asia-Pacific region). National authorities now have access to recently published international standards to guide
Agent Oriented Systems, ATFM, ATM, the development of national and regional operational concepts for ATFM, geared to Communications,
AMAN, ASBU, Capacity, CDM, Civil- Navigation, Surveillance/Air Traffic Management and Avionics (CNS+A) evolutions. This paper critically
Military Co-Ordination, CNS+A, DAM, reviews the field to determine which ATFM research and development efforts hold the best promise for
DCB, DMAN, Flight Object, FUA, PBN, practical technological implementations, offering clear benefits both in terms of enhanced safety and efficiency
PDS, PNT, Self-Separation, Sequencing, in times of growing air traffic. An evolutionary approach is adopted starting from an ontology of current ATFM
Situational Awareness, Slots, SMAN, techniques and proceeding to identify the technological and regulatory evolutions required in the future CNS+A
SWIM, TBO, UDPP, UPR, Weather context, as the aviation industry moves forward with a clearer understanding of emerging operational needs, the
Avoidance
geo-political realities of regional collaboration and the impending needs of global harmonization.

1. Introduction International Air Transport Association (IATA), the Civil Air

Navigation Services Organisation (CANSO), the United States
Air Traffic Flow Management (ATFM) involves a number of
(US) Federal Aviation Administration (FAA) and the European
measures to accomplish the mission of supporting a safe, efficient
Organisation for the Safety of Air Navigation (Eurocontrol).
and expedited flow of air traffic. Both long-term and short-term
measures are considered to resolve perturbations arising due to The US Next Generation Air Transportation System (NextGen)
unpredicted weather and capacity disruptions. The effectiveness and European Union (EU) Single European Sky ATM Research
of these measures depends considerably on the amount, accuracy (SESAR) research programs include significant ATFM
and timeliness of the information exchanged. As a result, human components. Recent ICAO publications (Doc 9971 - Manual on
operators depend crucially on technology enablers and Decision Collaborative ATFM, Doc 9965 Manual on Flight and Flow -
Support Systems (DSS) for making better informed and more Information for a Collaborative Environment, Doc 9854 – Global
effective decisions. Air Traffic Management Operational Concept, etc.) indicate a
legislative will to harmonise global ATFM and Collaborative
ATFM has existed as a component of Air Traffic Management
Decision Making (CDM) operational concepts and procedures by
(ATM) for several decades, supported to different degrees by
building on the experience of existing practitioners and expanding
largely bespoke automated systems in a handful of countries
their envelope of applicability [1]-[3].
(USA, European Union, Japan, Thailand, South Africa, Australia,
New Zealand, Mexico, Columbia and Brazil). Broader harmonisation initiatives build on ICAO’s 2008
NextGen-SESAR Gap Analysis and various Standards
In recent years ATFM has become pertinent even in regions
Roundtable discussions since. Developed with industry input, the
without sustained overload conditions caused by dense traffic
ICAO Aviation System Block Upgrade (ASBU) modules now
operations. Increasing traffic volumes in the face of constrained
form the unifying framework for the evolution of the Global
resources has created peak congestion at specific locations and
ATM System and it is in this context that any industry-relevant
times in many areas of the world. Increased environmental
discussion of ATFM/CDM should occur.
awareness and economic drivers have combined to create a
resurgent interest in ATFM as evidenced by the growing number For industry these developments mark the maturing of user
of recent ATFM initiatives mediated by official bodies such as the requirements in an emerging ATFM market segment. For
International Civil Aviation Organisation (ICAO), the researchers it is a prudent juncture at which to review recent
developments in the field of ATFM/CDM and to categorise them
_________
* Corresponding authors:
E‐mail address: [email protected] (T. Kistan), [email protected] (R. Sabatini),

http://dx.doi.org/10.1016/j.paerosci.XXXX.XXX
2 T. Kistan et al. / Progress in Aerospace Sciences

by industrial criteria such as technology readiness levels and strategy locally or invoked in the pre-tactical or tactical phase of
operational tenability. operations in order to compensate for failings in that strategy or in
This article includes specific sections discussing the fundamental its upstream implementation. ATFM can therefore be seen as a
operational settings: airport, approach and enroute operations, as subset of ATM, with some blurring of distinctions as
well as a section dedicated to regional ATFM measures and technological advances (in trajectory prediction, load estimation,
development initiatives. System Wide Information Management (SWIM) etc.) progress
towards the implementation of gate-to-gate concepts.
In terms of mathematical modelling, this article reviews the
fundamental problems, the key mathematical models and the With reference to the academic literature, various ICAO, SESAR
solution approaches that have been explored as part of several and NextGen industry documents, including the minutes of recent
ATFM research and development programs undertaken in recent ICAO ATFM workshops, established ATFM operational practice
years. Five aspects are discussed in particular: the sequencing (predominantly FAA and Eurocontrol) and vendor-supplied
problem, the arrival-departure problem, the gate-to-gate problem, product information it is possible to construct an up-to-date
as well as dynamic models of traffic flow and airspace capacity. ontology of ATFM techniques.
Global solution approaches are explored to promote optimised Since there are several operational improvement imperatives with
flows across a sizable network of airports and regional airspace. overlapping focus areas the most expedient approach is to classify
2. Ontology of ATFM techniques the ATFM techniques by phase of flight (Fig. 2). Indeed, the
networked nature of ATFM means that tactical measures applied
The continuous growth in air traffic demand underscores the locally at one point of the network are perceived to be pre-tactical
limitations of conventional ATM and ATFM procedures and measures by actors at another location in the network. For
operational paradigms. Novel operational concepts, technology comparison, a more traditional categorisation of ATFM
enablers and DSS are being developed to accommodate this techniques by time and separation criterion (vertical, horizontal,
growth trend as part of extraordinary aviation modernisation longitudinal or time) is given in the ICAO Manual on
initiatives currently taking place around the globe [100]-[106]. Collaborative ATFM [1]:II-6-2.
The new systems expected to emerge from these initiatives shall
One acknowledged consequence of departing from the traditional
address the multiple superimposed problems that characterise the
categorisation of ATFM techniques into planning/strategic, pre-
operation of airports and airspace sectors in the highly
tactical, tactical and post-operations analysis phases is that the
interconnected and interdependent air route network.
ontology in (Fig. 2) is missing the identification of the post-
Perturbations or disruptions occurring locally due to a variety of
operations analysis to planning loop that is meant to gauge the
reasons can quickly affect a growing number of airports, airspace
efficacy, efficiency and equitability of past ATFM measures and
sectors and flights unless effective mitigation measures are put in
lead to their improvement [1]:II-2-3.
place. ATFM is specifically tasked to continuously match the air
traffic demand with the available capacity of airspace and airport
Regional ATFM
resources. A number of ATFM reaction measures have been
developed over the years, addressing both the strategic (i.e. long- - Major Traffic Flows/LOAs
term) and pre-tactical / tactical (short-term) timeframes. - IATA Slot Scheduling
- Seamless ATM
ATFM measures can be modelled as mathematical optimisation - Open Skies
problems in which either an individual solution or a set thereof is
to be identified. Optimal solutions are those where the model
National ATFM
resembles the real scenario and all assumptions are valid.
Uncertainties in input data and estimation models necessarily void - Legislative Framework/Business Rules
some key assumptions, therefore typical solutions are pseudo- - Infrastructure/Capacity Increases
- Airspace Structure/Flextracks/UPR
optimal.
- Strategic & Pre-Tactical Capacity/Load Planning
Uncertainties in input data originate from the limited amount and - CDM Conferences/Web Portals
timeliness of information shared amongst flight crews, ATM - Daily Airspace Usage Plans
operators, Air Navigation Service Providers (ANSP) and Airline
Operations Centres (AOC). Typical ATM Operations
Airport capacity is often variable and dependent on weather
conditions, traffic composition, infrastructure conditions and
Aerodrome Operations Approach Enroute Approach Aerodrome
human factors. Meteorological factors that significantly impact Operations
capacity include wind direction and intensity, precipitation, - Arrival Sequencing - Arrival Seq - Wx Deviations ...
visibility, hazardous clouds and more severe hazards. The - Departure Sequencing - Departure Seq - Re-routing ...
variability introduced by meteorological factors is exacerbated by - Airport Slots - FUA/DAM - Exit Separation
- Ground Holds - Holding - FUA/DAM
the fact that aircraft have different weather capabilities ranging - Airport CDM - Vectoring
from almost all-weather to completely inhibited from operating. - Surface Management
The resulting airport capacity, in terms of arrival and departure Gate to Gate
rates, can vary considerably and abruptly.
Fig. 1. Relationships between ATFM and ATM.
Fig. 1 illustrates that while certain techniques can be considered
unique to ATFM, many of the sample shown are typical ATM
techniques that are applied either to implement a broader ATFM
3 T. Kistan et al. / Progress in Aerospace Sciences

Airport Approach Enroute Regional

Queue Management Queue Management Queue Management

Queue Management
- Time-Based Metering - Time-Based Metering -Time-Based Metering
- Pre-Departure Sequencing - Extended AMAN
- Arrival Sequencing - Waypoint Slot Manage
- Departure Sequencing
- Extended AMAN - “XMAN”
- Arrival Sequencing
- Point Merge - Exit Separation Manage
- Surface Management - Gates Dynamic Airspace Management
- Departure Sequencing

- Flexible Use of Airspace Seamless ATM

Trajectory-Based Operations
Capacity Management - Civil/Military Co-ordination
- Dynamic Sectorisation - Compatible Procedures
- Continuous Descent Arrival
- Time Based Separation
- Tailored Arrivals
- Wake Turbulence & RECAT Optimised Trajectories Optimised Trajectories
- Continuous Climb Departures

Capacity/Load Balancing - Enhanced Enroute Trajectories - Enhanced Enroute Trajectories

Capacity/Load Balancing - User Preferred Routing - Green Routes
- Arrival/Dep Rate Balancing - Trajectory Based Operations - Flextracks
- Fix Balancing - Traffic Synchronisation - Weather Avoidance
- Runway Balancing
- Airborne Holding - Weather Avoidance
- Surface Management - Parking

Slot Management Capacity/Load Balancing Capacity/Load Balancing

- Ration-By-Schedule Allocation - Fix Balancing - Declared Capacities

- Compression, Clustering - Airborne Holding - Letters of Agreement
- Ground Delays - Miles-in-Trail
- Ground Stops - Minutes-in-Trail
- Rerouting Regional CDM
- Level Capping
Airport CDM - Conditional Routes/”Playbook” - Teleconferences
- Interoperability
- SWIM
- SWIM
- Slot Swapping
- Datalink
- Data Exchange
- Virtual Centres

Fig. 2. Emerging Ontology of ATFM Techniques.

3. Airport and to the Regional Supplementary Procedures (Doc. 7030) for

the specific provision of separation minima based on prevailing
Airports frequently represent logistical bottlenecks and a source
circumstances [20]. Minimum separation distances and intervals
of unpredictability in the air route network. Consequently several
both on approach and on departure are dictated by the most
aspects of ATFM have a dependency on airport-related factors
restrictive requirements in terms of longitudinal separation.
and conversely, a number of ATFM-related measures are being
Whenever surveillance information is available to the ATCO, the
more deeply integrated into the operational paradigms for
minimum lateral (horizontal) separation distance is prescribed in
aerodrome control duties.
relation to the surveillance systems performance. Currently, the
The fundamental factor to be considered when modelling and standard minimum lateral separation distance in nautical miles
managing traffic flows is the acceptance rate, which qualifies the (NM), as recommended by ICAO based on either Primary
throughput of each element in the network. For airports, Surveillance Radar (PSR), Secondary Surveillance Radar (SSR),
throughput has traditionally been expressed in terms of aircraft Automatic Dependent Surveillance-Broadcast (ADS-B) and/or
movements per reference unit of time (i.e. hourly, daily, weekly, Multi-Lateration (MLAT) systems is 5.0 NM [21]. According
monthly or yearly), also referred to as to ICAO, the minimum lateral separation distance can be reduced
hourly/daily/weekly/monthly/yearly movement capacity. Since to 3.0 NM whenever the surveillance performance available
airports resort to operational reconfigurations and restrictions so permits, and a further reduction to 2.5 NM is finally
during the course of the day to better accommodate prevailing granted for aircraft established on the same final approach track
traffic flows and mitigate aircraft noise perceived by the resident within 10.0 NM from the runway threshold, provided that
population, hourly capacities are most commonly quoted. approach speeds, runway occupancy, braking action and wake
The hourly capacity of an aerodrome in terms of aircraft turbulence factors are known and continuously monitored by the
movements depends both on the number and layout of runways Air Traffic Controller (ATCO) [21]. This further reduced
and their related navigation facilities as well as on a multitude of minimum separation can be achieved only when a runway
additional factors associated with other ground infrastructure, occupancy time consistently below 50s is statistically
local geographic and meteorological characteristics. However, demonstrated and when PSR and SSR surveillance with an update
estimations based only on the runway layout do provide valuable rate of at least once every 5s is available with coverage reaching
reference figures, especially in terms of maximum capacity. up to and including the short final [21]. This surveillance
performance is normally provided only by approach-grade PSR
3.1. Arrival throughput for a single runway and SSR located within or in close proximity to the airport. En-
route grade PSR and SSR are typically not adequate, as the update
Annex 11 of the Chicago Convention on Civil Aviation refers to rate is reduced to achieve longer detection ranges and coverage at
the Procedures for Air Navigation Services-ATM (PANS-ATM)
4 T. Kistan et al. / Progress in Aerospace Sciences

low altitudes is limited to minimise ground clutter and/or due to departure traffic as this alternation obviates the wake turbulence-
the radar being located farther away from the airport. related longitudinal separation requirements. Consequently, when
Wake turbulence separation standards are conventionally defined arrival and departure routings are segregated, the total hourly
by distance (nautical miles) on approach and by time (minutes) on throughput in the case of mixed arrivals and departures is higher
departure, as these are the metrics most readily monitored by than in the case of arrivals-only or departures-only. This is
ATCO respectively in approach control and in aerodrome control represented conceptually by the Pareto plot of Fig. 3. In this plot,
tower duties. In particular, wake turbulence separation criteria the points A0 and D0 lying on the Cartesian axes represent arrival-
were defined by introducing 4 distinct categories based on the only and departure only operations. Points A1 and D1 represent
reference Maximum Take-Off Weight (MTOW) of aircraft: light traffic distributions where the residual hourly capacity, which
(up to 7,000 kg), medium, heavy (more than 136,000 kg but not could not be exploited in arrival-only or departure-only
including AIRBUS A380) and super-heavy (currently only the configurations due to wake turbulence requirements, is recovered
AIRBUS A380). by “squeezing in” some departures or arrivals. Point B represents
a “balanced” configuration.
The minimum longitudinal separation distances on approach can
be conveniently translated into separation intervals, , , by
considering the ground speed of the leading aircraft, , the

Departures Throughput
Single Runway
ground speed of the trailing aircraft, , the occupancy time of the
leading aircraft on landing, , the prescribed minimum
longitudinal separation distance between the leading and the
D0 D1
trailing aircraft due to wake turbulence, , and the minimum
lateral (radar) separation, . Two alternative cases are obtained, B
depending on the relationship between the ground speed of the
leading and trailing aircraft [23]:
,
, max , when 1 Feasible
Region A1
,
, , when 2 A0

Airports where surveillance systems and/or precision approach Arrivals Throughput

navigation facilities are limited or not available can implement
Fig. 3. Total hourly arrival and departure throughputs in the
analogous runway throughput estimation methods, though the
case of single runway operations.
more restrictive procedural minimum separation criteria will be
predominant.
In order to accommodate traffic throughputs beyond the
As mentioned before, additional factors affecting arrival and theoretical limit related to a single runway, and mitigate some of
departure throughput at a particular airport include ground-based the effects associated with adverse weather, airports resort to
infrastructure as well as the geographic and meteorological operating multiple runways simultaneously. The increase in
characteristics of the airport region. In particular, the close throughput associated with additional runways depends
proximity of mountain ranges or prominent peaks, as well as the fundamentally on their orientation and distance from the existing
frequent occurrence of low visibility, precipitation and/or strong runways, as ground intersections and under-separated flight paths
winds not oriented along the runways are some of the factors that will compromise the capacity increase. The total throughput for
can considerably reduce an airport’s maximum hourly the arrivals only and departures-only modes is nonetheless
movements. In terms of ground-based infrastructure, the increased whenever a better distribution by wake turbulence
movements can be limited in relation to taxiway and apron category across the available runways can be pursued. The
layouts, number of parking stands and capacity of airport services presence of multiple runways increases the number of possible
including turnaround and refuelling vehicles and personnel, as operational configurations, therefore in comparison with the
well as security and fire-fighting services. single-runway case the Pareto front for a multi-runway airport
Due to the considerable operational impacts associated with wake (Fig. 4) is normally characterised by a higher number of points.
turbulence-related longitudinal separation criteria in the Terminal
Manoeuvring Area (TMA), recent initiatives and ongoing studies
are addressing the possibilities to reduce the longitudinal spacing
Departures Throughput

while safeguarding the level of safety. In this respect, RECAT-2

and RECAT-3 are initiatives to maximise the benefit from
RECAT-EU – the European Re-Categorisation of wake
turbulence categories [87].
RECAT-2 defines pair-wise separations, allowing runway
throughput to be maximised by selecting an optimal traffic mix.
RECAT-3 goes further by proposing the dynamic adjustment of
pair-wise separations using real-time measurements of weather, Feasible
wake vortices etc. derived from both airborne and ground systems Region
[87].

3.2. Mixed mode and multiple runway operations Arrivals Throughput

Applying the minimum separation criteria on approach and
Fig. 4. Total hourly arrival and departure throughputs in the
departure can lead to inefficient runway usage, especially when
case of multiple runways.
the mix and sequence of traffic is suboptimal and/or highly
heterogeneous. In order to eliminate these sequence-related
inefficiencies airports traditionally resort to mixing arrival and
5 T. Kistan et al. / Progress in Aerospace Sciences

approach is often not viable as airports are gradually engulfed by

3.3. Airport capacity and flow management urbanisation that prevents successive expansions. Furthermore,
In order to operate at, or close to, the maximum hourly capacities, some bottlenecks cannot be removed and this can negate any
a precise synchronisation of all ground and runway operations is benefit offered by infrastructure expansion. The progressive
necessary. Consequently, disruptions to terminal and/or ground introduction of Communication Navigation Surveillance (CNS)
operations, which could easily span several minutes, can rapidly equipment and DSS with prediction functionalities is supporting
propagate not only to the entire airport but also to traffic within an evolutionary change to airport capacity and flow management.
the TMA and progressively beyond. In order to mitigate these Table 1, adapted from seminal works in the domain, outlines five
unavoidable perturbations and prevent or resolve major progressive levels of ATFM decentralisation that were conceived
disruptions, the historical strategy was to design in an operational in relation to the various ATM/ATFM tasks [108].
over-capacity at airports. In addition to being costly, this

Table 1. ATM/ATFM task implications of progressive levels of ATFM decentralisation [108].

Context Assignment of Arrival Assignment of TMA, Ground

Allocation of Arrival En-route Planning
Slots to Individual Departure Slots to Movement Planning
Service Slots and Control
Flights Individual Flights and Control

ATFM service assigns

slots, each airline ATFM service defines
Centralised
may cancel and the authorized routings.
substitute flights
ATFM service allocates ATFM service assigns
arrival slots to departure slots to
individual flights Each airline suggests individual aircraft
alternative assignment
of the slots allocated to
Partially Centralised
its own flights. ATFM
service approves or
ATFM service
rejects
prescribes and monitors
the routing as part of
the clearances.
Individual airlines
Partially Distributed allocated their own sets
(basic) of slots among their
own flights Airlines assign
ATFM service
ATFM service allocates departure slots to
delegates the routing
sets of arrival slots to individual aircraft,
within specific regions
individual airlines subject to ATFM
of airspace to airline.
Airlines may trade service approval.
Partially Distributed slots. Each airline
(advanced) allocates its own set of
slots among its flights

Airlines may bargain

Airlines plan the
ATFM service informs for the legal capacities.
preferred departure and
Distributed airlines of the legal They may confirm,
arrival routings, ATFM
(basic) capacities at potentially cancel or delay flights
service verifies and
congested airports within their purchased
Airlines assign approves or rejects.
slots.
departure slots to
individual aircraft at
Airlines plan and enact
their own discretion. Airlines plan and
ATFM service informs Airlines have full their preferred arrival
control their routings.
Distributed airlines about freedom in managing and departure routings,
ATFM monitors for
(advanced) anticipated availability the legal capacity for ATFM service
unavailability and
at congested airports. arrivals. monitors for feasibility
conflicts.
and conflicts.

proposed in the literature are presented here and throughout the

3.3.1. Scheduling problem rest of the paper.
A number of models that were developed in the area of operations Scheduling (or metering) problems arise whenever a resource is
research have been exploited to address the fundamental Demand subject to an air traffic demand greater than the available
Capacity Balancing (DCB) problems arising in ATFM. As a capacity. Solution methods normally attempt to determine the
counterpoint to the emergence of machine learning techniques resource allocation schedule that maximises some figure related
allowing to build models from data rather than from first to the efficiency in exploiting the available capacity, or
principles, some of the formal mathematical models that were conversely the one that minimises a total weighted delay. This
6 T. Kistan et al. / Progress in Aerospace Sciences

problem is equivalent to the single-machine scheduling with amendment of pre-departure clearances in case of changing
ready times of operations research. weather conditions. Recent works in this context are [109]-[113].
In the airport context congestion problems can arise due to:
3.3.4. Ground stop and ground delay programs
 a mismatch between the nominal capacity figures adopted
when granting slots to airlines and aircraft operators and the Ground Delays Programs (GDP) are some of the most traditional
actual capacity available when accounting for weather measures considered in the framework of ATFM. Originally these
conditions and other factors restricting the nominal capacity. Traffic Flow Management Initiatives (TFMI) were enacted
whenever a particular airport was experiencing unplanned
 the unexpected advance or delay of inbound traffic due to
reductions in capacity (typically due to adverse weather). These
factors shortening or extending the en-route time, such as
conventional GDP consisted of delaying the departure of any
winds aloft, hazardous weather cells or clear air turbulence,
aircraft headed to the affected airport from all departure airports
which can result in peaks of over-demand;
within a certain range from it. Traffic already en-route or
 other factors prompting inbound traffic to arrive earlier or
departing airports beyond the specified range would not be
later than planned, also including cascaded phenomena;
affected [97]. In its most severe form, GDP also involves the on-
 unscheduled “pop-up” traffic. ground holding of long-haul flights already in the air at
The problem can be stated as: for each inbound aircraft determine intermediate stop-over airports. The high uncertainty and
the Allocated Time (AT) that optimises a predefined performance variability of forecast data favour the application of GDP to short-
index considering the geometry of the approach, the Preferential haul traffic due to the limited flight time making short-term GDP
Time (PT) of each aircraft, the estimated runway occupancy and possible. This approach would theoretically make GDP more
the minimum longitudinal safe separation constraints. Higher efficient and effective as the accuracy of shorter-term forecasts is
degrees of complexity are introduced in relation to the differences higher and the more degrees of freedom are available for short-
in approach speed and wake turbulence category of various haul traffic. Nevertheless, fair scheduling principles require
aircraft, as well as to weather conditions which, as mentioned, can delays to be redistributed to all traffic in an equitable way.
vary the available capacity quite significantly. Therefore, GDP management models should account
stochastically for the uncertainties in forecast data and this can
3.3.2. Mathematical formulation and solution methods result in either more conservative strategies, where considerable
margins in the DCB are preserved, or more liberal approaches
The scheduling cost functional can be written in a generic form where the exploitation of all the predicted capacity is attempted.
that includes fictitious members both at the machine start time Traditional GDP also did not account for the potential disruptive
and between every job as [108]: effects to departure airports, and this aspect is increasingly
important as more and more major airports are operating close to
saturation conditions.
Δ Δ Δ (3)
Improvements to GDP practices are being considered in terms of
shifts in Controlled Times of Departure (CTD) to better match the
where S is the desired schedule, z is the performance index to be Controlled Time of Arrival (CTA), as this is the parameter
minimised, w are weightings and are delays. Fictitious jobs effectively dictating the magnitude of the delays to be enacted to
redistribute demand peaks with the aim of rebalancing demand
( ) are defined to run between the completion time of the and capacity at destination airports [97]. Additionally, GDP
previous job ( ) and the start time of the following ( ). This implementations that are not restricted in range and that also
generalised form allows tracking and, if desirable, assigning a concurrently consider Airspace Flow Programs (AFP) measures
cost to both natural and inserted machine idle times (modelled as are being investigated as they would contribute to enhancing the
the fictitious jobs) between different jobs. efficiency and effectiveness of these TFMI. An AFP is essentially
a GDP applied to meter traffic transiting through a Flow
The first possible solution approach consists in prioritising the
Constrained Area (FCA) of airspace. An AFP is resorted to when
ready job for which the ratio between the job duration and the
short-term en-route measures (such as miles-in-trail and re-routes)
penalty weighting ( / is minimal. In this approach, which can
prove insufficient to regulate the traffic flow rate through the
be classified as a Simple Cheapest Insertion Heuristic (SCIH)
FCA.
[108], no idle time is inserted intentionally as the job to be
inserted must be selected from the ready queue. This scheduling
method, called Heuristic H and documented in [114] is shown to
require log steps. A more sophisticated method, proposed 4. Approach
by Belouadah and belonging to the family of known lower bounds As discussed in section 3 the hourly movement capacity of an
methods, considers the interruption of a running job if the airport is the key parameter defining its aircraft acceptance rate.
duration time-to-weighting ratio ( / of a newly ready job From the TMA operations perspective the acceptance rate is a
compares favourably with the one of the already ongoing job metering constraint set on the ensemble of airport runways
[114]. The ongoing job is split between the already completed allocated to arrivals at any given time. The acceptance rate of a
portion and the remaining portion. particular airport normally corresponds to the arrival throughput
of active runways. In relation to the acceptance rate, further
3.3.3. Surface management metering constraints are set upstream (i.e. backwards) along all
Improved surface operations, an objective of the US NextGen the arrival routes, normally defined at merge points set further
R&D initiative, aims to ensure the automatic reassessment and out, and in the case of very busy major airports, up to enroute
update of the departure schedule thanks to improved data sectors farther than 150 NM away. Other operational factors
exchange, better surface movement surveillance and more including workload limitations for aerodrome and/or TMA ATM
effective management measures. The updated information is operators might reduce the theoretical capacities and may pose
expected to enable a more efficient and effective scheduling of further limitations to the acceptance rates.
runways, metering points and ATFM variables. Data-link According to ICAO, the capacity of an Air Traffic Services
communications are to be used to streamline the dispatch and (ATS) unit depends on (3.1.1.1 and 3.1.2) [21]:
7 T. Kistan et al. / Progress in Aerospace Sciences

 the level and type of ATS provided; Where is the MAP capacity G is the workload intensity, T
is the transit time of the sector, is the mean hand-over service
 the complexity of the sector/area/aerodrome and of the time. A workload intensity of 1 equates to the capacity limit in
associated route structure; nominal conditions. The mean hand-over service time was found
 ATCO workload, including control and coordination to be currently set to
tasks to be performed.
36 (5)
 navigation accuracy of aircraft and CNS system
performance available; The transit time for a particular sector can be either computed
 availability of support and alert systems; from statistical data or with an empirical estimation such as:

 weather; √
(6)
The ATS authority is responsible for assessing and declaring the ̅
capacity of control areas, control sectors and aerodromes [21].
Capacity should be expressed as the maximum number of aircraft Where A is the horizontal surface area of the sector, and ̅ is
which can be accepted over a given period of time in the the average True Air Speed (TAS) of the traffic in the sector.
concerned area [21] (3.1.1.3), and this most often translates into
hourly traffic flow limits. Some recommendations for the An allowance of plus or minus three aircraft was also identified.
assessment and definition of ATS capacities were provided in Consistent with the assumptions of the MAP model a maximum
Doc. 9426 Air Traffic Services Planning Manual[25]. The method safe throughput of one flight every 36 s, which equates to 100
developed by the United Kingdom Directorate of Operational aircraft per hour, was determined for sectors with transit times up
Research and Analysis (DORATASK) and the method developed to approximately 11 minutes. Above that figure the throughput
by the German Messerschmidt Bolkow und Blohm (MBB) have decreases according to the limitation of 18-aircraft within the
been extensively used to estimate the effort required for the sector. The authors of the study identified that peak flow
management of individual traffic and to determine safe total capacities of up to 120 aircraft per hour were statistically
hourly capacities for each sector [25]. Both methods can be occurring, but only for sectors with transit times of 300
applied to TMA and to enroute sectors. The DORATASK method 600. As a consequence, sectors with volumes between 5000 and
distinguished between the time allocated to observable tasks, such 30000 NM3 were identified having the highest flow capacities,
as deconfliction, and the time spent on non-observable tasks such but only if a higher density of the air route network was planned
as planning and periods of recuperation. Time spent on for increasingly smaller sectors, since the maximum flow along
observable tasks is recorded and used to develop simulations for individual routes is constrained by the longitudinal separation
accurate estimation. Time spent on non-observable tasks is requirements.
empirically postulated based on some measurable parameter such Considering these factors, the authors of the study determined that
as the number of flight strip updates. The total ATCO workload is the MAP model consistently over-estimates the safe capacity of
calculated as a sum of time spent on observable and non- smaller sectors, and did not account for the flow directionality.
observable tasks, plus a percentage reserved for recuperation.
This percentage was initially modelled to increase along the work An enhanced model was therefore proposed in which the total
shift to account for the cumulative effect of uncertainties and to workload intensity consists of four aggregate components,
compensate for ATCO’s fatigue. Upon model validation it was including a “background” monitoring component , a recurring
ultimately found that the average workload should have been task component, a hand-over component and a deconfliction
limited to 80% on a continuous basis and that peak loads of up to component as follows [98]:
90% should not have occurred more than 2.5% of the total time
[25]. The MBB method involves the identification of working 1
units associated with working actions and the identification of an (7)
airspace-related factor defining the conflict probability [25].
Welch et al. [98][99] investigated the current operational capacity Where 2 , is the mean service time associated
model used by the FAA in the National Airspace System (NAS) with recurring tasks, P is the associated recurrence period, is
and based on the identified limitations developed a promising the mean service time associated with deconfliction tasks, and
improvement. The current model, called Monitor Alert Parameter are the horizontal and vertical distances associated with
(MAP) is used by the FAA to trigger operational traffic alerts, separation violations, is the mean closing speed of aircraft
mainly by monitoring the hand-over traffic amounts. This that incur separation violation and Q is the airspace sector volume
simplified approach accurately estimates the inter-sector [99]. The dependency on the fraction of aircraft that are either
coordination workload and its linear growth with traffic count. climbing or descending more than 2000 ft ( ) within the sector
The higher traffic densities typical of smaller sectors, leading to can also be integrated [98]. A further modification of this model
higher conflict rates, are effectively considered as the theoretical to account for weather impacts was proposed [98]:
capacity that is linearly proportional to the transit time. As a
consequence, within the NAS sectors with transit times lower 1
than 4 minutes were found to be fairly rare, as their maximum (8)
1
capacities would be up to 7. On the other hand, recurring tasks
grow linearly with traffic amount. Consequently in the MAP
This formulation includes terms associated with the weather-
model the capacity limit for large sectors is no longer dominated
induced mean service times for both recurring ( ) and hand-
by the linear dependence on transit time but by a hard limit count
over coordination tasks ( ) multiplied by the sector’s weather
(18 aircraft). The current model can be written as [98]:
blockage factor .

min ; 18 (4)
8 T. Kistan et al. / Progress in Aerospace Sciences

Aerodrome Traffic Zone (ATZ). Four aircraft of different

4.1. Arrival sequencing make/model are approaching the airport, characterised by an
The sequencing problem for arrival traffic is conceptually individual runway. The segment in white represents the common
represented in Fig. 5, where two different arrivals route are path where minimum safe longitudinal separation must be
depicted across the TMA, the Control Area (CTA) and the ensured at all times.

TMA

CTA

ATZ

AOC ANSP ATM

GROUND COMMUNICATION NETWORK

(COLLABORATIVE DECISION MAKING)

Fig. 5. Conceptual representation of the arrival sequencing scenario.

aircraft performance or aircraft dynamics model which accurately

4.1.1. Arrival managers estimates the variations of airspeed, flight path angle and turn
The purpose of Arrival Managers (AMAN) is to ensure an radius across altitudes. The trajectory prediction functionality
optimal sequencing and spacing of arrival traffic, therefore requires inputs of surveillance data, weather data, operational
extending the implementation of ATFM concepts and measures restrictions and aircraft performance. The demand is determined
beyond the en-route airspace to encompass arrival routes and the initially by considering unconstrained crossing times at
TMA. AMAN are crucial in minimising the need to resort to low- predefined metering points. The sequencing is calculated based
level vectoring and holding patterns, which lead to high fuel on these nominal crossing times and also considering other
consumption, perceived noise on the ground, exhaust emissions important factors, primarily the minimisation of inefficiencies
and negatively impacts on passenger comfort. This is done by caused by sequences of traffic with highly heterogeneous wake
ensuring the necessary spacing between arrivals is achieved as turbulence categories.
early as possible. If a delay due to unpredicted perturbations In order to generate the advisories to human operators
arises AMAN allows it to be absorbed in the most efficient surveillance data, weather data, flight plan information, wake
manner, for instance by distributing it linearly along the entire turbulence categories, nominal aircraft performances, local route
residual route to the destination. AMAN systems assist human and airspace structures are combined to generate accurate
controllers in performing their sequencing and spacing duties and trajectory predictions, normally by means of the Flight Data
feature dedicated functions such as the computation of Estimated Processing System (FDPS) and the Radar Data Processing System
Time of Arrival (ETA) at future route fixes in real-time, (RDPS). Constraints, either predefined or set by the operator,
verification of desired traffic separation up to the Final Approach include throughput and minimum longitudinal and lateral
Fix (FAF), time to lose and time to gain in the sequence, optimal separations. The AMAN operational horizon refers to the area
runway allocation for arrivals and slot allocations for departures where advisories are generated. An eligibility horizon can be
in mixed-use runways. The optimisation of arrival scheduling and defined to include all traffic for which Advanced Boundary
determination of required delays are accomplished by Information (ABI) is available [59]. A sequencing process is
implementing a trajectory prediction and a sequencing element. executed to determine the ABI of traffic in the operational
The trajectory prediction functionality must implement a suitable horizon for which no ABI is supplied.
9 T. Kistan et al. / Progress in Aerospace Sciences

All automated advisories can be customised according to different available will be maximised. To construct a departure sequence at
user-defined goals including minimum total delay, minimum runways and SID (Standard Instrument Departure) exit points a
average delay or minimum deviation from preferred profiles [59]. DMAN will typically:
ETA and Requested Time Over (RTO) are visualised in suitably
defined formats both on the active trajectory overlay within the  Be intimately involved in Airport Collaborative
tactical display and in a dedicated distance-time visualisation. Decision Making and receive information from CDM
partners (such as ATFM, ATC, Advanced Surface
Recommended speeds to achieve optimal spacing are also
Movement Guidance and Control System (A-SMGCS)
provided to the ATCO in dedicated visual formats. Advanced
DSS are capable of suggesting the recommended action for the and Airline Operators).
ATCO to consider from a wide array that includes path stretching  Compute optimal gate & runway threshold times taking
and holding patterns. into account known delays to scheduled off-block times,
AMAN systems are crucially important to mitigate disruptions to variable taxi times, runway acceptance rates & runway
the arrival throughput caused by unpredicted events including arrival sequences for mixed mode operations. An off-
adverse weather and runway closures. AMAN efficiency is block time is the time that the aircraft pushes back from
maximised when an integration with Departure Managers the gate.
(DMAN) is implemented, as the optimised slot allocation can be  Optimise a combination of taxi time, runway utilisation
extended to encompass departures. This is discussed further in
or conformance-to-schedule criteria taking into account
section 4.3.
engine start-up times, push-back procedures, line-up
times, separation criteria etc.
4.1.2. Extended AMAN
 Utilise gaps in the arrival sequence to maximise runway
Eurocontrol has developed the AMAN Information Extension to usage.
En-route Sectors – Concept of Operations [26] which is aligned
 Suggest a departure runway, runway access holding
with the SESAR ATM Maserplan [12] supporting the Queue
Management Tools operational capability and laying the points and a SID.
foundations to the First ATM Support Tools Implementation  Continuously update the sequence on a time-ladder or
(FASTI) programme [26]. This document defines the pathway for on an integrated electronic strip display & allow the
AMAN systems and concepts to be extended beyond the TMA. controller to manually modify the sequence, insert
This is crucially important in the European region as a large reservation slots & model “what-if” scenarios.
number of major airports are located in proximity to each other
and to Flight Information Region (FIR) boundaries, and also  Perform conformance monitoring of departures against
because the airspace jurisdiction is highly fragmented. The ATFM slot windows. A slot window is a permissible
concept is promising for regional ATFM outside of Europe as tolerance around the off-block time (e.g. -5 to +10
well. minutes) before which operational or commercial
As part of SESAR ongoing research is investigating systematic penalties may apply.
methodologies for resilience management in ATM [27]. The  Improve the predictability of the trajectories used by
resilience management problem is mathematically formulated as ATC and ATFM systems.
an optimal task reallocation problem. This investigation is
providing valuable insights in the perspective of evolving ATFM A DMAN may also offer the capability to:
models and technologies as the effects of significantly higher
levels of automation needs to be modelled in future DSS.  Consider restrictions at departure fixes.
 Manage flow rates at SID exit points, supporting
4.1.3. Long range metering / pre-sequencing departure splitting.
 Consider runway dependencies at other managed
Pre-sequencing prior to the AMAN Feeder Fix occurs in enroute
airspace. In the Asia-Pacific context this can occur within enroute airports (e.g. the use of a common departure fix).
sectors that are adjacent to neighbouring FIRs and impact the  Consider ATFM constraints and AMAN advisories from
Letters of Agreement (LOA) governing exit separation criteria. remote departure aerodromes.
CDM techniques will be required if these criteria are to be  Consider flight-specific prioritisation rules.
modified in real-time. An alternate approach is to treat the Co-
Ordination Point (COP) as a metering fix and sequence flights Current research foresees the DMAN being cognisant of airport-
through it (Exit Separation Management). In this respect, and airline-related CDM constraints, as well as operational
Australia has been discussing Long Range ATFM concepts with runway, TMA and regional control centre constraints. With
Indonesia and the Philippines [69]. respect to airport-related CDM constraints, DMAN evolutions
should accommodate limitations associated with de-icing,
4.2. Departure sequencing baggage handling, boarding delays and taxi times. In terms of
airline CDM constraints, DMAN evolutions should consider
A DMAN aims to optimise airport departure sequences in the face
individual targets/priorities and the availability of datalink
of multiple constraints. In particular, DMAN automatically
services. In terms of operational constraints, DMAN evolutions
compute and monitor the required separation and determine the
should consider runway closures, online computation of wake-
optimal sequence for departures from individual, intersecting or
vortex separations, TMA exit separation criteria, miles/minutes in
closely spaced runways. In addition to minimum spacing due to
trail requirements, departure splitting, regional ATFM slots and
wake turbulence criteria, the DMAN decision support algorithms
Traffic Flow Management Initiatives (TFMI) as well as
take into account the specific departure routings of successive
downstream sector congestions or acceptance rate variations.
aircraft. The goal of DMAN is to prevent queues of departing
aircraft to form on the ground, and hence considerably reduce fuel
consumption and noxious emissions. DMAN will benefit from A-
CDM and SWIM as both the amount and timeliness of input data
10 T. Kistan et al. / Progress in Aerospace Sciences

airports the DMAN and AMAN co-ordinate departures at one

4.3. Integrated arrival and departure sequencing airport with arrivals at the other. For closely-spaced arrival
Contemporary research topics in arrival/departure sequencing airports the AMAN sequences several airports in the TMA
leverage the improved trajectory prediction anticipated with the through common or different feeder fixes. For closely-spaced
use of Flight Management System (FMS) data, Aircraft Derived departure airports with common exit fixes, DMAN systems can
Data (ADD) for intent and performance data. New procedures, synchronise with each other to achieve a common departure
such as Continuous Descent Arrivals, allow for the downlink of stream. For farther-spaced city pairs (national or international) an
an accurate time window for ETA on future trajectories, allowing (eXtended AMAN (XMAN) can serve as an intermediary
an AMAN to compute a Controlled Time of Overflight (CTO) between AMAN and DMAN systems.
that can be uplinked to the aircraft before it arrives at its Top of
Descent. 4.3.2. Departure-arrival problem
System Wide Information Management enables more detailed Departure-arrival problems involve more complex scheduling
sequencing advisories, for example: problems arising when the resources are shared by different
categories of users i.e. departures and arrivals. An optimal
 Actual sequencing actions (holding, Standard Arrival alternated allocation must be determined. The departure/arrival
(STAR) or speed advisories) to be undertaken in problems can be formulated as multi-machine scheduling
specified enroute and approach sectors, named holding problems on parallel unrelated machines [108]. These, in turn, can
patterns etc. be approximated as assignment problems in order to be solved in
 Consideration of delays already issued by a DMAN or polynomial time.
Slot Manager at the aerodrome of departure to avoid
penalising a flight twice. 4.3.3. Mathematical formulation and solution methods
 Automated consideration of curfews and noise As mentioned, the departure-arrival problem can be formulated as
abatement procedures. a multi-machine scheduling problem on parallel unrelated
 Integrated ATM-AMAN/DMAN “what-if” modelling machines. The multi-machine scheduling model at the basis of
this formulation must assume that [108]:
features
 Multiple optimisation criteria (slot compliance, runway  Each machine represents a single airspace/airport resource;
utilisation, sequence stability, taxi delays, equitable  Each aircraft involves either a single job or no job at every
treatment of airline operators, …) machine;
 The job ready time represents the PT of a particular aircraft at
The AMAN/DMAN planning horizon is extending outwards to a particular airspace/airport resource in its route;
several hours allowing for consolidated arrival, departure and
 The job starting time represents the AT of a particular aircraft
runway usage management in a multi-TMA configuration that
at a particular airspace/airport resource in its route;
treats airports as nodes in the ATFM network. In this scenario,
indications of early intent / departure planning information  The job processing time represents the time required by an
become increasingly important; as does real-time data from ATM, individual aircraft to clear/vacate the corresponding resource.
A-SMGCS, Multi-lateration etc. The scheduling cost functional can be written as
Shared AMAN/DMAN runway configuration and data
management will enable the balancing of arrival & departure rates
in accordance with local runway utilisation, availability of , , Δ , , , , (9)
parking bays and planned airframe turnaround times.
AMAN/DMAN integration will support mixed-mode runway
operations. One approach is to extend runway acceptance rates Where , , represents the cost of processing the jth job on the ith
with ratios between arrivals and departures (e.g. 4 arrival and 2 machine using lth slot. Assuming a constant processing time ,
departures). The AMAN builds the sequence, leaving gaps for X for any job on any machine the multi-machine scheduling on
departures for DMAN to use. This facilitates integrating an parallel unrelated machines can be approximated as an
AMAN from one vendor and a DMAN from another (SESAR assignment problem [116]. This assignment problem can be
features an AMAN from Thales and a DMAN from SAAB). represented by the bipartite graph of Fig. 7. Each job concerns a
Another approach is time-based AMAN/DMAN co-operation, different subset of machines and each machine has a different
illustrated in Fig. 6. number of slots.
The solution of this assignment problem corresponds to the
solution of the original problem subject to the condition that all
job durations are equal. A solution to the assignment problem is
shown to be achievable in √ log steps.

Fig. 6. Time-Based AMAN/DMAN co-operation

4.3.1. Multiple airport arrival and departure management

Integrated AMAN/DMAN systems can support multi-nodal
ATFM as follows. For closely-spaced arrival and departure
11 T. Kistan et al. / Progress in Aerospace Sciences

MACHINE 1
SLOT 1

SLOT 2
JOB 1
SLOT 3

JOB 2 MACHINE 2
SLOT 1
SLOT 2
SLOT 3
JOB 2
SLOT 4

JOB 2
⁞
Fig. 8. Representation of arrival flows encompassing two
MACHINE m alternative fixes.
SLOT 1
⁞ SLOT 2
4.4.2. Airborne holding
SLOT 3
Airborne holdings are ATM/ATFM measures conceived to either
SLOT 4 delay an aircraft or temporarily stop its progression along its route
JOB n
or to the next Air Traffic Control (ATC) sector. These measures
may become necessary in relation to either a traffic demand
exceeding the acceptance rate of the destination airport, or when a
Fig. 7. Bipartite graph representation of the multi-machine downstream ATC sector has already attained or exceeded its safe
scheduling on parallel unrelated machines. capacity and cannot accept additional traffic. Airborne holdings
have rather severe impacts in terms of fuel consumption,
especially when the holding takes place at low altitude.
4.4. ATFM techniques for the approach context
Consequently most ANSPs pursued the development of en-route
The dense route and airspace structure of the TMA and the ATFM measures as well as airport/route network/airspace
relatively short timeframe characterising approach and departure redesign initiatives aimed at minimising the occurrence of
operations significantly restricts the number and extent of ATFM airborne holdings.
initiatives that can be pursued to improve traffic flow optimality.
A number of conventional and more innovative techniques are 4.4.3. Point merge
nonetheless available to mitigate the effects of perturbations.
Point merge is a relatively new arrival sequencing technique
designed by the Eurocontrol Experimental Centre to replace radar
4.4.1. Fix balancing
vectoring [61]. It utilises a common merge point and several pre-
Fix balancing involves the repartition of traffic flows across defined legs, each equidistant from the merge point. The legs are
multiple gates. This technique can involve both arrival and used for path stretching or shortening in order to smooth peak
departure flows and is implemented on a regular basis at a load (simultaneous arrivals) upstream of the individual legs [60].
number of airports as it delivers consistent benefits . The main The serial combination of several point merge systems (for
advantages associated with fix balancing include the repartition of example, to reflect TMA constraints) has been contemplated but
traffic flows depending on the arrival/departing runway and not investigated [61]:80.
maximising the flow capacities on the routes by segregating
different aircraft categories to reduce the negative effects 4.4.4. Trajectory Based Operations (TBO)
associated with alternating wake turbulence category or
heterogeneous traffic. Additionally, geographically separated In the Trajectory-Based Operations (TBO) operational
traffic flows can be handled by different ATCO thus relieving paradigmeach aircraft is associated with a 4-Dimensional
workload limitations on the flow capacities. Fig. 8 depicts arrival Trajectory (i.e. in terms of latitude, longitude, altitude and time)
routes at a single runway airport. It can be seen that arrival flows which is negotiated, monitored and updated continuously along
from the south and the north can be distributed amongst two the route. This represents an evolutionary change with respect to
alternative fixes. Cross-links (represented in blue and violet) are legacy flight plans. TBO are expected to improve considerably
also available. the arrival and departure sequencing and spacing functions and
are considered an essential tool to progressively migrate the ATM
system from its current largely centralised and tactical nature to a
more strategic and distributed paradigm.

4.4.5. Miles-in-trail
Miles-in-trail is a technique that specifies the distance-based
longitudinal separation required to achieve a particular flow rate
along a route allowing for various factors including separation
12 T. Kistan et al. / Progress in Aerospace Sciences

standards, spacing required to accommodate merging traffic circumstances TFMI either in terms of GDP or AFP are required
[1]:II-6-1 etc. to redistribute the demand to a more uniform profile, as depicted
in Fig. 11. An example of the Human-Machine Interface and
4.4.6. Minutes-in-trail Interaction (HMI2) of a current state-of-the-art ATFM DSS is
represented in Fig. 12. A single sector is considered and the
Minutes-in-trail serves a similar purpose to miles-in-trail and is demand is sampled every 20 minutes. The capacity is depicted in
employed where distance-based separation is difficult to achieve red, and an instance of over-capacity, highlighted in yellow, is
(such as low-surveillance environments) [1]:II-6-1. noticed in correspondence with a temporary capacity reduction.

4.4.7. Re-routing
Tactical re-routing can be applied for a number of ATFM Demand
purposes – weather diversion, sector load balancing, path Capacity

Demand and Capacity [units/h]
Under‐capacity
stretching in order to establish a particular sequence etc. Over‐capacity

4.4.8. Level Capping

Level capping applies flight level restrictions flights to avoid
congestion until the flight exits the congested zone (the TMA, for
example [1]:II-6-2).

4.4.9. Minimum Departure Intervals

A minimum departure interval throttles back the departure rate of 3 6 9 12
Time [h]
15 18 21

an aerodrome to avoid temporary congestion in a departure sector

[1]:II-6-2 or at a Standard Instrument Departure exit point. Fig. 9. Relationship between original demand and
capacity in an ACC.
4.4.10. Slot Swapping
Slot swapping is an aspect of Airport CDM that allows airline
operators to swap their airport slots (with other airline operators if Demand
Capacity
required) in order to better meet their business objectives [1]:II-6-
Demand and Capacity [units/h]

Under‐capacity
2). This reduces the network impact of operational delays that the Over‐capacity

airline operator has already suffered.

5. Enroute
Demand-Capacity Balancing (DCB) is a key objective of ATFM
that is applicable across all three ATFM phases and which seeks
to regulate the flow of traffic such that demand matches available 3 6 9 12 15 18 21
capacity. DCB is most useful when imbalances are predicted early Time [h]
enough to avoid tactical ATC intervention. If this is the case then
TFMI performed in the enroute context are the most effective. Fig. 10. Representation of ASM initiatives in terms of sector
consolidations and de-consolidations (green arrows).
Consideration of a sufficiently useful time horizon for effective
DCB underscores the fact that ATFM benefits from a regional,
cross-border approach. The two fundamental approaches that are
considered, typically in synergy, to pursue DCB in the enroute Original Demand

context are AFP and Airspace Management (ASM). A number of Amended Demand
Capacity
Demand and Capacity [units/h]

evolutions were considered for both approaches and are currently Under‐capacity
being developed as part of all major ATM modernisation Over‐capacity

initiatives. Challenges to the effectiveness of ATFM measures in

the enroute context arise from uncertainties in trajectory
prediction and weather forecasts over the longer ranges involved.
Fig. 9 conceptually represents the relation between the predicted
traffic demand and the planned airspace capacity for an Area
Control Centre (ACC) facility.
As depicted by the step shifts of the blue line in the figure, the 3 6 9 12 15 18 21
Time [h]
capacity incurs discrete changes as airspace sectors are
consolidated and de-consolidated and as special use airspace is Fig. 11. Simultaneous changes to demand and capacity due to the
reserved and released. In terms of ASM, an optimised timing of concurrent implementation of AFP and ASM initiatives.
these consolidations and de-consolidations can be implemented to
help minimise the under-capacity instances by optimally fitting
the predicted demand. This is represented by Fig. 10. An overall
maximum ACC capacity associated with the available workforce,
CNS/ATM infrastructure and airspace geography may pose a hard
limit for which sector de-consolidations may not be sufficient on
their own to prevent under-capacity situations. This is
exemplified by the morning traffic peak in Fig. 9. In these
13 T. Kistan et al. / Progress in Aerospace Sciences

authorities may prefer to run just under their maximum capacity

to afford themselves the flexibility for opportunistic ATFM
actions and tactical intervention. The need to reserve some
residual capacity may be greater where the business rules
governing ATFM measures have not been institutionalised and
unscheduled traffic may be anticipated.
Following the ICAO manual on collaborative ATFM [1], it can be
noted that DCB can be implemented for resources such as
airports, airways, fixes or sectors by employing specific ATFM
measures across all phases of operation. Strategic DCB can be
addressed by measures such as IATA airline schedule planning,
airport slot co-ordination, ATM planning, airspace and procedure
design, capacity analysis and performance prediction. Pre-tactical
DCB measures include weekly and daily CDM meetings and
Flexible Use of Airspace planning while tactical DCB measures
include:
 Exit Separation Management and Miles/Minutes-in-trail
techniques.
 Fix balancing and movable feeder fixes.
 Ad-hoc and pre-defined reroutes (conditional/
preferential/alternate/“playbooks”).
 Level capping and tunnelling.
2
Fig. 12. HMI of THALES TopSky ATFM software, depicting an over-  Minimum departure intervals.
capacity situation (used with permission).
 Ground delay and ground stops.
International initiatives seek to harmonise the methodologies by  Airborne delay absorption and holding.
which different authorities determine and publish deemed
capacities for resources. For example, sector capacity can be Traditional ATM automation systems provide some level of
specified by occupancy (the maximum number of aircraft that are support for implementing ATFM measures, particularly during
simultaneously under the jurisdiction of the controller managing the tactical phase of operations. Emerging ATFM DSS provide
the ATC sector at any time) or by flow rate (the maximum additional capabilities, including:
number of aircraft that can enter that particular ATC sector in a  Demand prediction based on scheduled, filed or live
specified period)[25]. Several authorities have considered or traffic data and/or analytical estimation/numerical
adopted capacity measures based on ATCO workload. In simulations [27].
principle such an approach has potential safety benefits; in
practice, while methods like DORATASK have been applied to  Fast-time simulation and “what-if” modelling of ATFM
determine workload-based capacities, no standardised equivalent measures.
method for computing workload-inspired dynamic-density
demand (current or forecast) has yet emerged. Workload is  Graphical visualisation of traffic (air situation display)
dependent on a complex interaction of several factors [25], many and load (bar charts or similar).
of which are difficult to predict in advance. Traffic complexity  Common situational awareness (display of weather,
permits a simplified estimation of ATCO workload. It seeks to aeronautical messages, constraints, etc.).
characterise the cognitive effort associated with a particular
configuration of airspace and air traffic. A number of  CDM communication mechanisms (increasingly web-
unpredictable factors can manifest during operations, including: based).

 Delays on departure.  Compliance reporting and post-event analysis.

 Queuing for take-off. Major areas of current research for ATFM DSS include HMI2
factors to visualise complexity and dynamic density concepts,
 Unscheduled, pop-up traffic. merging metering with sequencing (i.e. the integration with
 Winds aloft differing significantly from the forecasts. AMAN/DMAN operations) and the development of regional
ATFM concepts that support multiple ANSPs. From the specific
 Weather deviations and avoidance of turbulence. DCB measures listed above, it can be observed that DCB
addresses predominately demand-side issues; the ability to
 Tactical ATC intervention for separation and sequencing. modify capacity-side factors is currently limited to strategic and
These factors introduce significant uncertainties into the predicted pre-tactical measures. This limitation is addressed by Dynamic
traffic demand, with some (like winds aloft) having the potential Airspace Management (DAM).
to impact not just individual flights but clustered flight flows due
to their geographic extent and relatively long duration. ATFM 5.1. Mathematical framework
DSS typically acknowledge these uncertainties by differentiating
In parallel with the modelling of human workload characterising
their representation of active and forecast traffic demand by
studies that built upon the DORATASK and similar approaches
confidence criteria, permitting flow managers to tolerate some
some early work attempted the mathematical modelling of
degree of forecast low-confidence overload if strategies exist to
airspace capacity based mainly on operational considerations
distribute it if it does indeed eventuate. Conversely, when there is
[29]-[31]. The more recent advances in Communication,
a high level of confidence in the forecast demand, some
14 T. Kistan et al. / Progress in Aerospace Sciences

Navigation, Surveillance, ATM (CNS+A) and Avionics By introducing constraints associated with congested airspace
technologies are progressively enabling enhanced DCB strategies. sectors in aircraft trajectory planning algorithms it is possible to
The new CNS+A technologies greatly increase the flexibility of numerically force trajectory planning across less densely
routes and airspace configurations, which are both of great use for trafficked sectors. As an example, Fig. 14 depicts a rerouting
DCB. Nevertheless, since at present ATM is still largely based on instance around congested airspace sectors of the Swiss Area
instructions dispatched verbally by air traffic controllers in Control Centre (ACC).
individual radio transmissions, workload is still clearly identified
as the constraint dictating airspace capacities [32]. Recent work
such as [33], [34] introduced more complex and accurate models
and airspace constructs. Some preliminary Human-In-The-Loop
(HITL) evaluations of the novel flexible airspace concepts led to
positive acceptance by ATM operators [35], but an identified key
factor was the adoption of automated transfer-of-communication
and more generally of TBO data-com equipment. ATFM and
conventional ASM algorithms consider sectors as entities with
several entry and exit points that handle traffic entering, exiting
and flowing through. This principle has been adopted by many
large scale models [33], [36]. Fig. 13depicts two airports O and D
where flight f is planned to depart from and arrive at, respectively.
In order to arrive to airport D, flight f must pass through either
sector V , V or V , which are characterised by various number of
entry and exit points. This permits flight f to access alternative
routes in case one of the original planned sectors becomes
unavailable due to severe weather, which is an example of a pre- Fig. 14. Flight planning considering a congested sector.
tactical or tactical ATFM initiative.
5.1.1. Demand prediction models
Traditional demand prediction methods rely on trajectory
estimation in open-loop. This is done by retrieving or estimating
the entry and exit times of each sector along the path.
A normalised weather index is typically adopted for each sector
representing the portion affected by severe weather.
Future Automation Concepts Evaluation Tool (FACET) evaluated
advanced ATFM simulation tools to forecast both future traffic
states and their related confidence levels.
Both traditional methods and FACET methods are open-loop and
thus do not take into account combined ATFM measures [117].
Fig. 13. Large-scale network model of airspace sectors [33].
Chen and Sridhar developed a prediction model that takes into
Each sector is represented as a volume of space, which can account ATFM actions as well as tactical ATM and airline
accommodate a number of aircraft depending on the throughput actions, and is suitable for midterm predictions [117]. A power
measured at the sector entry/exit points. The traffic either enters law distribution can model the demand reduction as:
from adjacent sectors or originates from an airport within the
sector itself. In order to define the network, each sector is
connected to other sectors or airports. The traffic flow can be 1 (11)
expressed in terms of aircraft entering and leaving the sector v
during each time interval t as in:
where w is the lower weather index threshold and w is the
weather index causing the entire sector capacity to be sequestered.
, , , ,
∈ ∈ 5.1.2. Macroscopic traffic flow modelling
(10)
, , As the implications arising from the discreteness of the traffic
∈ ∈ diminish with the increasing size of the volume considered,
continuity can be at some point assumed when pursuing a
macroscopic modelling of the traffic demand across very large
The number of aircraft in the sector, N , , is equivalent to the areas. In this respect, the Lighthill-Whitham model, modified by
number of aircraft previously in the sector N , , and to the Richards to account for first-order continuity – the so called LWR
number of aircraft entering and leaving the sector at each entry model – was the first macroscopic traffic flow model based on the
point e ∈ and exit point x ∈ X. The variable k here represents continuity principle to be proposed [118]. Several researchers
any capacity metering element such as airports or sector entry/exit have developed Eulerian network models applicable to air traffic
points. An overall capacity constraint can be defined on the total flows based on modified versions of the LWR model [127]-[135].
number of aircraft simultaneously in the sector: N , N . The LWR model is based both on the traffic densities, ρ(x,t),
When a sector is affected by severe weather, a capacity reduction defined as number of aircraft per unit length along route
is dynamically introduced as the weather cell progressively segments, and on traffic fluxes, q(x,t), defined as the throughput
impacts portions of the sector volume in terms of altitude and in number of aircraft per unit of time along route segments. The
horizontal extents. Capacity constraints for airspace sectors following continuity equation expresses the LWR model:
require flights to reroute around most congested sectors.
15 T. Kistan et al. / Progress in Aerospace Sciences

deviations from the planned arrival schedules at each and every

0 (12) airport/termination node as:

where s is a source term that accounts for aircraft either entering (25)
(s>0) or exiting the sector (s<0) not from peripheral nodes. This
source term can be omitted by introducing opportune virtual
nodes/segments where the fluxes are prescribed and this is where , is the number of arriving aircraft
desirable from the numerical implementation perspective. In at the end of link kα.
particular, Work and Bayen [127] reviewed various previous
works [128]-[135] and improved the LWR-based problem The work also assesses a few numerical solution strategies for the
formulation to achieve a strictly convex implementation, ensuring resulting hyperbolic problem based on finite differences and
global optimality. For a network of links k, the problem is complying with the Courant-Friedrichs-Lewy (CFL) condition for
formulated as: numerical stability. To ensure that problem convexity is retained
in the numerical implementation, particular consideration is given
Minimise: to the discretisation of boundary conditions.
Eulerian network flow models based on LWR continuity support
, , , , , (13) the assessment of the demand and capacity of various airspace
resources (either of airway route segments and/or of airspace
sectors and regions), which are of great value for airspace/air
s.t.: route network design and their assessment in terms of resilience
to disruption. These models are therefore particularly promising
, , in the implementation of DAM. On the other hand, in addition to
0 ∀ (14)
being valid only at macroscale (i.e., regional or continental scale),
these models do not capture the specific departure and destination
0 , , , ∀ (15) needs of each individual aircraft, therefore scheduling algorithms
from operations research are needed to simulate the actual state of
, , , , , ∀ (16) the network, evaluate the downstream impacts of every TFMI or
contingent capacity restriction and perform the ultimate decisions
,0 ∀ (17) in terms of clearances dispatched to aircraft.
,

5.1.3. Gate-to-gate problem formulation and solution methods

0, , , , ∀ (18)
Gate-to-gate problems involve a global formulation in which
departure scheduling, arrival scheduling and flight routing are
, , , , ⋅ , ∀ (19) considered together to attempt maximising the overall system
efficiency. This problem can be translated to a multi-commodity
, , , , ⋅ , ∀ (20) flow problem [108]. The generalised formulation of this problem
involves a set of airports A interconnected to any other in the set
0 ∀ (21) through an air route network Λ, consisting of nodes and arcs.
Each node is characterised by a distinctive 3-dimensional position
0 (22) in space. Each arc can only connect two nodes and is
characterised by a specific geographical distance. Airports are
where , is the portion of flights exiting link m that will enter
also modelled by their associated ground route networks, which
link k, so ∑ 1. Eq. 13 is the cost function, where J is the include a number of nodes associated with gates, runways and
,
performance index. Eq. 14 are the dynamic constraints ground intersections, which may or may not be connected by
introducing continuity along each and every link. Eq. 15 and 16 taxiway arcs. Within a time period T, a set of aircraft K has to
are linear inequality constraints on each and every link. Eq. 17 are leave a gate of a departure airport and reach a gate of the
initial conditions. Eq. 18 are boundary and interface conditions. In destination airport. The ready time of a specific job represents the
terms of the cost function, the work proposes three alternatives all PT of that particular aircraft at a particular node. The start time,
based on fluxes. The first possibility is to maximise the flux of conversely, represents the AT of a particular aircraft at a
arrivals at each and every airport/termination node as: particular node. Each aircraft can fly each segment at
predetermined minimum and maximum speeds, respectively
and . As a consequence, an upper limit is introduced to the
, (23) delay accumulated en-route.

The gate-to-gate problem was formulated as a multi-machine

Alternatively, it is possible to minimise the deviations from the scheduling problem in stages by Guercio [108]. By defining , ,
original flight plans (which must be known in terms of densities the length of arc ai,k,j connecting machine i at stage j with machine
and fluxes on each and every link):
k at stage j+1, , , the speed of job jl on the arc ai,k,j connecting
machine i at stage j with machine k at stage j+1, we can
(24) determine the maximum arc delay as:

where and are respectively the desired fluxes and densities. 1 1

Dependencies on space and time were omitted for succinctness. , , , , (26)
The third alternative proposed involves the minimisation of the , , , ,
16 T. Kistan et al. / Progress in Aerospace Sciences

By defining ,
the transit time of any job on machine i at stage A number of heuristic solution strategies were proposed. A
heuristic algorithm following the First-Come First-Served (FCFS)
j, and with , , the time spent by job l on arc ai,k,j we have the
principle at the destination involves the following steps:
following relationships [108]:
1. Generate an ordered list of aircraft by increasing RT at a
particular destination (if airborne) or by their PT if still
, , (27)
, , on the ground;

, ,
2. Assign the first available and attainable slot to the first
, , , (28) aircraft in the list by solving the shortest path problem;
, ,

3. Delete the first aircraft and the assigned slot from the
, ,
, , , , , (29) list and repeat steps 2 and 3.
, ,
An improved version that attempts the solution on en-route
, , conflicts is:
, , , , , (30)
, , 1. For each aircraft ordered by increasing RT at a
particular destination, assign the first available slot

, ,
(mimicking the conventional FCFS heuristic algorithm
, , , , , (31) above);
, ,

2. For each aircraft which transits a particular machine

, ,
, , , (32) within the same slot already used by another aircraft,
, , assign the first successive slot in that machine;

3. Repeat step 2 for all aircraft and all machines until all
The problem can be solved in terms of , ∀ ∀ by introducing slots are deconflicted.
suitable constraints to ensure the continuity of the path of each
job, that minimum and maximum speeds (and associated arc
delays) are met, that job processing on any machine starts at the 5.2. Dynamic Demand Capacity Balancing
beginning of an available slot, that any two jobs do not share the
same slot on the same machine, that the processing of one job on Consideration of specific techniques reveals that Capacity
a specific machine does not start before the ready time and finally Demand Balancing addresses mainly demand-side issues (Table
that the start and finish of each job coincide with the desired 2). Flexible Use of Airspace techniques permit the temporary
start/finish machines. The gate-to-gate problem presented above relief of capacity restrictions by modifying the timing or
can be translated to a multi-commodity flow problem in a geographic extent of reserved Special Use Airspace – they cannot
particular directed graph. In this formulation a set of commodities create “new” capacity. Emerging DAM concepts introduce a true
share common arc capacities that restrict the total flow of tactical supply-side factor and are described in the following
commodities on that arc. A viable solution algorithm for this section. Next generation ATM systems are expected to support
implementation involves the following steps [108]: the technique which means that ATFM systems should include
the concept in their roadmaps. In a regional context, co-ordination
issues have to be addressed if sectors adjacent to the FIR
1. Prune the initial set of nodes of the ones not involved in
boundary are modified, but the technique could prove to be an
any aircraft path;
alternative to permanent airspace restructure and calls for FIR
consolidation.
2. For each aircraft find the minimum and maximum time
required to reach each vertex from the departure vertex;

3. Based on the determined minimum and maximum times Table 2. Some Capacity-Demand Balancing Techniques.
find vertex time intervals usable by all aircraft;
Strategic Pre-Tactical Tactical
4. For each vertex, based on the transit time, find the IATA Airline Daily Airspace Exit Separation
admissible slots included in the time intervals. For each Scheduling; Usage Plan; Management;
slot, the capacity of the vertex-slot is 1; Airport Slot Co- CDM Meetings; Miles/Minutes-in-
ordination; Flexible Use of trail techniques;
ATM Planning; Airspace; Fix balancing,
5. For each admissible couple of vertex-slots establish in Airspace and Re-Sectorisation movable feeder fixes;
the network the associated arc; Procedure Design; Ad-hoc and pre-
Capacity Analysis; defined reroutes;
6. For each aircraft, assign for each arc of the network a Performance Level capping and
pseudo-infinite cost if the arc may not be used, or a cost Prediction tunnelling;
proportional to the distance plus a cost proportional to Minimum Departure
the delay; Intervals;
Ground delay,
7. Insert fictitious vertexes in relation to the departure and ground stops;
destination vertexes to model the supply and demand. Airborne delay
absorption, holding
17 T. Kistan et al. / Progress in Aerospace Sciences

AB1 decrease it and the aim of DCB in relation to SUA is to ensure

AB4
that the timing of that decrease in capacity coincides with a
AB2
decrease in traffic demand.

AB8 The flexible/dynamic airspace concept was introduced to

overcome the rigidity of this conventional airspace structure.
DAM concepts allow for a temporary increase in ATC sector
capacity by modifying the geographic extent of the sector itself,
AB5
morphing its boundaries and those of its neighbours to
accommodate shifting traffic patterns, moving weather or other
AB3 dynamic factors [15]. The capacity of the neighbouring ATC
sectors may be temporarily decreased as a result, but this is a far
AB9
more flexible concept that consolidating or de-consolidating
sectors onto ATC positions. Regulatory considerations (ATCO
ratings, traffic mix, etc.) may constrain the extent to which the
AB6 sector boundaries may change but relatively moderate
AB7
perturbations may still yield significant operational benefits for
both ANSPs (maximise ATCO performance and manage staffing
Fig. 15. Dynamic Sectorisation (adapted from [47]). levels) and airline operators (minimise DCB measures such as
reroutes and holding).

5.3. Dynamic Airspace Management Experimental studies on Flexible Airspace Management (FAM),
an early dynamic sectorisation investigated in NextGen,
In parallel with the implementation of 4D-TBO, the ATM highlighted the critical aspects in terms of the magnitude,
research community is tackling the concurrent evolution of the frequency, timing and lead time preview of sector changes [27].
airspace into a dynamically-optimised resource, taking into In particular, shifts involving large traffic volumes, major traffic
account traffic, airport and weather updates and forecasts. The flows or reverting upstream/downstream directions were found to
long-term evolution of ATFM and airspace management is have significant negative implications for both the workload and
represented by the DAM paradigm. While ATFM measures deal situational awareness of the ATCO.
largely with the modulation of the demand, DAM is conceived to
alter the capacity, by overcoming the rigidity of this conventional It was also realised that the current practice involving Letters of
airspace structure. Although regulatory considerations and Agreement (LOA) to regulate co-ordination at international
international agreements restrict the potential DAM measures, the boundaries posed a hard constraint for DAM and only allowed
increasing levels of information sharing and interoperability rigid and predefined rearrangements. It is nonetheless envisaged
amongst future ATM systems will enable a progressive relaxation that more data sharing and interoperability amongst ATM systems
of the constraints. A number of different automated partitioning will enable a progressive relaxation of cross-border arrangements.
techniques have been evaluated for future DAM DSS Significant research is required on the implications for inter-
implementations [37]-[39]. sector and inter-centre co-ordination.

Although differences are evident amongst various national DAM techniques should not be restricted to ATC sectors and
regulatory frameworks, three distinct regulatory aspects SUA. In the long term, ATS service classes, SUA and ATC
characterise the airspace structure: sectors should be coherently modulated based on the inputs of
suitable DAM DSS, adopting integrally dynamic algorithms for
 ATS classification: introduces a partitioning of airspace that is estimation and decision logic based on the available CNS
essentially based on the levels-of-service provided by the performance level. In this perspective DAM is considered to be
ANSP. ATS classes are classified from Alpha (A) to Golf (G) an essential contributor to PBO.
in decreasing levels-of-service with different requirements for
airborne equipage associated with each ATS class. The 5.4. Advanced dynamic sectorisation
classification carries significant implications. For example,
distinct rules of the air are applicable for each class. The feasibility and opportunity of automated partitioning
techniques has been discussed for a considerable time and some
 Airspace Restrictions/Special Use Airspace (SUA): are potential implementations were numerically evaluated in [35],
portions of the airspace characterised by a significant access [37]-[54]. In a possible Dynamic Airspace Sectorisation (DAS)
restriction or by a natural or manmade hazard to aircraft process, the conditions of weather and airspace segregations are
operations. SUA are frequently associated with segregated evaluated at each time period to identify the relinquished airspace
operations conducted by military, government, industrial or portions. Subsequently, the current configurations of airspace
scientific research organisations. SUA may also be designated sectors and traffic flows within the ACC jurisdiction are
to protect major cities or strategic and sensitive facilities from processed. A check on the balance between the estimated demand
criminal threat. SUA may be defined pre-tactically or and the updated capacity is therefore performed. In case an under-
tactically (i.e. online in real-time) and have several capacity or over-capacity is identified load balancing and ASM
activation/de-activation times. resolutions are attempted, for instance in terms of sector
 ATC sectors: reflect the organisational needs of the ANSP to consolidations and de-consolidations in combination with AFP
provide services to large traffic volumes over a vast resolutions such as re-routing traffic across less congested sectors.
geographic area. Airspace is partitioned into smaller sections If the demand-to-capacity balance is still suboptimal, further
that are under the jurisdiction of individual ATCO. attempts are undertaken.

Under traditional Flexible Use of Airspace (FUA) concepts the By introducing dynamic expressions of capacity and
geographic extent and activation times of SUA may be changed to demand for the i-th airspace sector, DCB is governed by the
minimise disruptions or maximise usage; however the ATC following expression:
sector(s) containing the SUA remains unchanged. SUA therefore
cannot increase the capacity of an ATC sector, it can only
18 T. Kistan et al. / Progress in Aerospace Sciences

The dependence upon time is introduced by taking into account

(33) the aircraft predicted to enter and exit the airspace sector as per
eq. 34. Consequently, the total sector capacity is also correlated
with the minimum guaranteed level of integrity provided by all
where is the sector occupancy trend. The term representing the CNS systems and therefore:
the variation in demand within airspace sector i can be expressed
as the balance between the rate of traffic entering in the sector, ∝ min , , (38)
the rate of traffic exiting and traffic originating or terminating
within the sector as:
The dependence on the minimum CNS performance levels
guaranteed by the entire sector traffic and not on the average is a
(34) result of assuming homogeneous traffic separation models in the
entire sector, which is a requirement of a significant increase in
the ATCO workload.
where represents the flow of demand (hence of traffic) from/to
the adjacent airspace sectors, and is a source/sink term DAS algorithms consider sectors as entities with several entry and
accounting for airports within the sector. In order to comply with exit points that handle traffic entering, exiting and flowing
an arbitrarily limited sector capacity it is possible to introduce through. This principle has been adopted by many large scale
thresholds for the predicted demand which will trigger a local models [33], [36]. Each sector is represented as a volume of
airspace reconfiguration. In particular, neglecting dynamically space, which can accommodate a number of aircraft depending on
morphing sectors, Flexible Airspace Management (FAM) the throughput of traffic at the sector entry/exit points. The entry
strategies involve sector splitting and merging. Research is traffic either enters by adjacent sectors or originates from an
currently underway to assess the viability of a variable sector airport within the sector itself. In order to define the network,
capacity and 4-Dimensionally morphing sectors [38]. each sector is connected to other sectors or airports. The traffic
flow can be expressed in terms of aircraft entering and leaving the
The key aspect lies in the introduction of capacity models that sector v during each time interval t as in equation 3. In line with
entail 4-Dimensional dynamics. We present an approach that is current airspace partitioning practices that rely on the 2D
based on the forthcoming CNS+A operational scenario. The novel representation of geographic boundaries plus a floor and ceiling,
CNS+A systems being developed for Performance Based the constraints typically introduced in DAS algorithms consist of
Operations (PBO) are conceived to meet the high CNS right prism shape for the sector volume and a convex polygon for
performance requirements that will enable an increase in the its geographic extent. Based on these typical requirements, the
traffic density without compromising flight safety. This can be partitioning of enroute airspace can be approached as a 2D plus
implemented by expressing the sector capacity as a function of time problem. Hence, out of the various algorithms proposed for
the CNS performance as follows. The capacity factor of airspace DAS, we present one example based on triangular tessellation and
sector i due to Communication Performance (CP) alone can be sequential consolidation. In particular, Fig. 16 depicts a medium
expressed as a factor of the minimum CP guaranteed by all resolution discretisation of an approximately 500 nmi x 250 nmi
aircraft j concurrently in the sector, as an example in terms of ACC into 34 triangular partitions by means of a 2D meshing
Communication Integrity (CI): algorithm. Fig. 17 depicts a higher resolution discretisation into
approximately 150 triangular partitions. As described in the
∝ min , (35) previous sections, the DAS/ATFM algorithm analyses the traffic
flow and attempts a consolidation/de-consolidation of the airspace
to optimise the demand-to-capacity balance over a certain time
Similar expressions hold true for Navigation Integrity (NI) and
horizon. The objective of 6 sectors is introduced to minimise the
Surveillance Integrity (SI):
required ATC workforce and attain high demand-to-capacity
optimality. An example of the algorithm result is represented in
∝ min , (36) Fig. 18 where the traffic flows are simplified by straight vectors
depicted in blue.
∝ min , (37)
19 T. Kistan et al. / Progress in Aerospace Sciences

Fig. 16. Discretisation of the ACC en-route airspace with a medium density mesh.

Fig. 17. Discretisation of the ACC en-route airspace with a denser mesh.
20 T. Kistan et al. / Progress in Aerospace Sciences

Fig. 18. Results of the DAS/ATFM algorithm.

In this case, the originally introduced objective of 6 sectors was Cooperative ATFM System (BOBCAT) system for the
not attained, as the algorithm identified that 8 sectors were management of waypoint slots in the Kabul FIR [71]. BOBCAT
required to maintain a sufficient capacity along the defined time implements long-range regional ATFM for westbound traffic
horizon of 3 hours. transiting Afghanistan airspace en-route to European destinations
[78]. BOBCAT went operational in July 2007 and currently
As the dynamics involved in DAS, load balancing and aircraft
manages airside entry gates into the Kabul FIR by issuing slot
trajectory planning are mutually interdependent and all are
allocations to aircraft at the departure airports [78]. BOBCAT
fundamentally based on real-time optimisation algorithms
works out an allocated wheels-up time from the gate entry
significant research is required to address the loop dependence,
window and the various fix, route and procedural constraints in
stability and optimality metrics, promoting rapid convergence to
place [78]. The regional nature of these measures is highlighted
global economic/environmental optimality. With respect to
by the fact that flights could be en-route for up to seven hours,
trajectory planning, in the short-term new airspace models will
transiting up to 10 FIRs with different separation standards before
have to consider dynamically evolving lateral and vertical sector
reaching the Kabul FIR [78]:35. BOBCAT uses a website to
boundaries and continuously varying capacities. In the medium-
implement the CDM aspect of ATFM. BOBCAT has been
long term, new combined traffic flow and airspace models
supplemented with an in-house ATFM Information Support
integrating DAS will be required.
System to better monitor regional load [71]. Thailand is actively
involved in ATFM information sharing with Singapore [72] and
6. Regional ATFM Initiatives Hong Kong, China and the development of a regional multi-nodal
ATFM concept [57]. Singapore has been involved in the
6.1. Asia Pacific development of the regional multi-nodal ATFM concept [57] and
is expected to procure a system soon. Hong Kong, China has
Asia Pacific is ICAO’s largest region, comprising 38 states and
supplemented its AMAN with an in-house Air Traffic Capacity
50 FIRs. The current Japanese ATFM centre was established in
Display System to better monitor load [70]. It has also developed
1994 and employs ATFM techniques such as level capping,
a CDM web portal [70] and is actively involved in regional
rerouting, metering (miles in trail and speed control), airborne
ATFM trials with Thailand and Singapore.
holding, ground stops, exit separation management, civil military
co-ordination, Flextracks and CDM [62]. NTT Data is currently Australia’s first attempt at ATFM was the Controlled Departure
contracted to replace the existing Japanese ATFM centre in time Time Program run at Sydney around 2000 [56]:4. However, the
for the 2020 Tokyo Olympic Games [63]. In 2014 India precursor to the current National Operations Centre (NOC) in
contracted Atech and Deutsche Flugsicherung (DFS) to provide Canberra was the Central Traffic Management System that was
an ATFM system [64]. After first focussing on Airport-CDM, housed in Sydney [56]:3-7. The software used for Slot
China is beginning to address both national and regional ATFM Management was an in-house developed Web application called
issues [65], starting with the Sanya FIR managing the South SKYFLOW that was quite sophisticated for its time. This has
China Sea [73]. The National Key Laboratory of Air Traffic Flow since been replaced by the Metron Harmony suite [67] and
Management at the Nanjing University of Aeronautics and supplemented by a nationwide CDM implementation [68]. The
Astronautics is tasked with researching organisational, NOC also runs in-house software ALOFT (ATM Long Range
operational and technological aspects of ATFM. Korea issued a Optimal Flow Tool) - a coarse long-range metering tool with a 4-
Request for Tender (RFT) for an ATFM system in 2015 with an hour look-ahead horizon that manages the early morning curfew
anticipated delivery date of 2017. Thailand continues to operate rush into Sydney airport [54]:15. This serves as a pre-sequencer
and update its long-standing in-house regional Bay of Bengal
21 T. Kistan et al. / Progress in Aerospace Sciences

for the Thales AMAN installed in the ATM system. The NOC is  Adaptive Compression: is a technique that identifies
also responsible for generating Flextracks using the FAA’s potentially unused slots and moves other flights into them,
Dynamic Oceanic Track System (DOTS).
minimising delays and maximising capacity. 
New Zealand has developed an in-house ATFM system - an
automated, Web application that manages ATC arrival slots [66]. Flow Management and CDM are key aspects of the FAA’s
Slot swapping is supported if airlines are not satisfied with the NextGen strategy. Version 2 of the NextGen Concept of
automated allocation [66]. A compression algorithm periodically Operations (CONOPS) addresses [76]:
optimises the slots and can be disabled for specific flights to
selectively freeze the metering for stability [66]. Uncertainties in  Collaboration for ATM, airport and airspace operations. 
the arrival schedule are managed by “clumping” arrivals around  Trajectory-Based airspace and operations, including super-
the same slot and colour coding is used to indicate the level of density arrival/departure operations. 
confidence in the data [66]. As a precaution, slots are reserved
around clumps and multiple missed approaches to prevent  Net-Centric infrastructure services supporting a shared
overbooking [66]. An interface to the ATC system supplies flight situational awareness and incorporating aeronautical,
progress data and allows slot time to be updated. weather and geospatial information services into decision
EurocontrolEurocontrol Central Flow Management Unit (CFMU) making, as well as flow strategy and trajectory impact
ATFM messages such as the Slot Allocation Message (SAM), analysis. 
Slot Revision Message (SRM), etc. can also be generated [66]. In
recent years the system has been integrated with the Exelis There has been some reprioritisation since [77].
Orthogon AMAN to achieve greater efficiencies in automation
[66].
6.3. Rest of the Americas
6.2. United States of America In preparation for the 2016 Olympics, Brazil contracted local
company Atech to modernise its ATFM system (deployed in
A combination of economic, regulatory and industrial factors in 2003) using its new Sky Flow product [85].
the 1980s gave rise to the first concrete implementation of CDM
in the United States. Industry deregulation and the adoption of Mexico deployed an in-house ATFM system called PROSAT in
hub and spoke operations by airline operators increased airport 2002 [84]. PROSAT is a basic ATC Slot Manager that works with
demand, while the oil crisis and safety concerns prompted by the flight plan data extracted from RPLs and AFTN messages. It has
1981 controller strike created a desire to reduce airborne holding implemented simplified versions of the EurocontrolEurocontrol
[75]:1. The FAA responded by implementing a program of CFMU ATFM messages for slot allocation, revision and
ground delays at departure airports to better balance capacity and confirmation known as “FLU”. A basic trajectory manager works
demand at arrival airports. Ground delay and airspace flow out arrival load at Mexico City and manages arrival slots by
programs assist traffic managers to distribute delays equitably assigning departure delay to feeder aerodromes. The algorithm
among impacted flights while maintaining the safe metering of works by considering the number and colour-coded confidence-
traffic. level of flights in 15-minute time blocks. In practice excess load
proves not to be an issue if there are a large number of low-
CDM capabilities are a key enabler of regional ATFM. A shared confidence flights as any surge can be managed tactically.
picture of factors such as airspace structure, reservations and Remotely deployed in-house software allows PROSAT to
usage plans, developing weather hazards, resource capacities, communicate with approach and tower controllers via the
traffic load and flight progress information permits flow managers proprietary “FLU” messages. Mexico City is IATA slot co-
to analyse traffic constraint problems and to develop and ordinated, but PROSAT cannot ensure ATC slot compliance with
communicate collaborative strategies for their resolution. the IATA commercial airport slots. Mexico has commenced the
Common strategies include ground delays and dynamic rerouting rollout of the new SMART ATFM centre and may procure a
of aircraft. Several CDM and ATFM tools are currently used by smarter Slot Manager in order to address this issue.
the FAA:
Columbia modernised its long-established ATFM centre with
 Flight Schedule Monitor: contributes to common situational Metron’s Harmony product in 2013 [86].
awareness amongst users and service providers making
them aware of constraints in the National Airspace System 6.4. Europe
in order to enable collaborative air traffic decisions. The
The Eurocontrol CFMU was created in the late 1980s in response
tool presents airport and airspace demand and capacity to the increasing traffic volumes and delays experienced during
information and is used to manage ground delay and that decade. By 1996 the CFMU was responsible for centralised
airspace flow programs. air traffic flow management over the member states of the
 Flight Schedule Analyser: analyses the efficiency of ground European Civil Aviation Conference (ECAC), thereby
consolidating the activities of 5 regional flow control centres. The
delay programs and the CDM process. 
CFMU is tasked with making the best possible use of airspace by
 Route Management Tool: a centralised route database balancing capacity and demand. The primary mechanism for
facilitates information exchange between air traffic control establishing this is the management of departure slot allocations
and the airline user community.  (the Computed Take-Off Time (CTOT)). However, Eurocontrol
strategic planning acknowledges that ATFM should be extended
 Traffic Management Advisor: helps controllers sequence
to the optimisation of traffic flow patterns and that capacity
aircraft through en-route airspace into major terminals by management should be achieved by collaborative decision
calculating fix crossing times in accordance with separation making rather than centralised control [74]. This process has
standards. Arrival streams can be established that take better already commenced with the implementation of FUA concepts
advantage of available landing slots.   and the Eurocontrol Airport CDM project.
The Eurocontrol Airport CDM project focuses on the efficient
and transparent sharing of work processes and data amongst
22 T. Kistan et al. / Progress in Aerospace Sciences

CDM partners, allowing each partner to optimise their decisions is the ability to visualise the broader network impacts of local
in collaboration with the others and with a common actions.
understanding of their preferences and constraints. Adoption of
Business/Mission Trajectory Management complements the
Eurocontrol Airport CDM concepts has commenced at many
conventional flight planning process with a Shared Business /
large European airports close to congestion.
Mission Trajectory (SBT) that is determined collaboratively and
Various European initiatives advocate current and future airport iteratively as new constraints and more accurate information
implementations such as: emerges closer to the time of flight.

 The establishment of collaborative rules defining the Scenario Management for Demand Capacity Balancing &
information to be shared and the events (milestones) used to Airspace Management pre-defines sets of basic DCB solutions
(e.g. airspace reconfiguration, temporary route structures, level
keep the CDM partners “synchronised”; 
capping etc.) for specific (and usually recurrent) demand capacity
 The deployment of an AMAN, a DMAN and a Surface imbalance issues. Response times are quick with predictable
Manager (SMAN) - a function of the A-SMGCS;  outcomes.
 Improved Airport-ATFM slot synchronisation;  User Driven Prioritisation Process (UDPP) is a collaborative
 Providing airports with an accurate estimated landing time process where stakeholders can recommend to the Network
up to three hours before, and then using an accurate take-off Manager a priority order for flights affected by delays caused by
an unexpected reduction of capacity.
estimate from the airport to update the flight plan for ATFM
purposes – including improved downstream landing Dynamic Demand Capacity Balancing affords users the freedom
estimates;  to change Business/Mission Trajectories until shortly before
departure and even during execution based on their operational
 Collaborative sharing of turn around information to improve needs. This requires “just-in-time” decision making supported by
estimates of subsequent events such as off-blocks and take- the Network Operations Plan.
off; 
A number of other SESAR work packages (5, 6, 10 and 12 [120])
 Improvements to the Airport Operational Plan (AOP) also deal with Flow Management and CDM issues such as:
impacting resource scheduling, local demand and capacity
balancing and queue management;  Co-operative Planning in the TMA concentrates on collaborative
planning for complex TMA environments serving multiple
 Improvements to runway throughput, including procedures airports. One focus will be on how different CDM processes and
to minimise the impact of adverse weather conditions.  support tools (such as AMAN, DMAN, and UDPP) can interact.

The current initiatives on Flow Management and CDM will be Ground and Airborne Capabilities to Implement Sequence
supported and accelerated in the coming years by the SESAR manages sequences by issuing and applying CTAs. CTAs smooth
Program, within which these subjects have particular emphasis. the distribution of arrivals via direct routing and improved
SESAR looks at airport operations from an "en-route to en-route" vertical profiles (including CDAs). A particular consideration is
perspective, with the aim of managing the aircraft turn around and the ability of automation systems (ATM, AMAN etc) to
flight operation as a single continuous process. Airports therefore synchronise traffic in a mixed equipage environment where not
need to be fully integrated into the ATM network. Broader ATFM all aircraft are Controlled Time of Arrival (CTA) / Required Time
has been recognised as a strategic topic within SESAR and of Arrival (RTA) capable.
warrants its own work packages under WP7 Network Operations Airport Capacity and Flow Management considers the
[120] to support trajectory-based operations, including: operational and technical requirements to interface airports to the
Integrated Network CDM is concerned with the integration of the Network Manager in order to harmonise the Airport and Network
Airport Operations Management (AOM) and DCB processes with Operations Plans.
improvements in enroute, terminal and airport operating contexts Integration of Airport–Airline/Ground Handlers – ATC Processes
contributing towards Network Operations. (incl. turn around) in ATM is an aspect of Airport Collaborative
Advanced Flexible Use of Airspace pursues the ad hoc delineation Decision Making that has a particular impact on Pre-Departure
of airspace structures across national borders and the procedures Sequencing and DMAN systems.
for temporary airspace reservations. Coupled AMAN-DMAN seeks to maximise the use of available
User Preferred Routing assesses route networks based on runway capacity by increasing the interaction between AMAN
business needs and dynamic airspace management concepts, and SMAN/DMAN systems. Runway delays can be reduced by
transitioning to user preferred-routing in areas of low to medium optimising inbound and outbound traffic flows.
traffic complexity. Integration of Queue Management seeks to develop an integrated
Dynamic Airspace Configuration performs a similar assessment set of queue management tools including AMAN, DMAN and
of en-route sectors and TMAs based on business needs and SMAN but extending to queue management in the en-route phase
dynamic airspace management concepts such as the real-time via techniques such as multiple CTO constraints.
modification of sector geometry and TMA boundaries to match Multiple Airport Arrival / Departure Management solutions for
changes in demand, user intentions and exceptional situations complex, multi-airport TMA environments, including, optimising
such as weather. Dynamic TMA configuration will be necessary delivery of traffic to multiple runways with minimal route or time
to cope with changes in allocated arrival/departure routes based extensions and enable constant throughput into the en-route phase
on traffic complexity. of flight. Extension of the AMAN time horizon into the en-route
Collaborative Network Operations Plan (NOP) is a virtual 4D phase.
model of the ATM environment that can be panned in space and Enhanced Sequencing Tools to handle departures from multiple
time. It enables SWIM-mediated global situational awareness of airports and flow interactions from adjacent airports.
dynamic capacity, resource availability, demand, forecast and
nowcast weather conditions, status of negotiations etc. Of interest
23 T. Kistan et al. / Progress in Aerospace Sciences

Integration of Departure Management and Surface Management overarching principles guide the development of a Regional
pursues the full integration of departure management, surface ATFM CONOPS from the perspective of Seamless ATM. The
management and arrival management, including the development implementation of a regional ATFM should be evolutionary
of a single supporting platform (e.g. a merged AMAN, DMAN rather than radical in nature, and should adopt a “systems
and/or SMAN). approach” capitalising on the ICAO ASBU as a roadmap towards
achieving seamless ATM in Asia-Pac. The new regional system
TMA and En-Route Co-operative Planning seeks to implement
should be compatible with the ATM operations of each authority
Collaborative Planning in TMA and En-route operations by
and hierarchical, in that local ATFM initiatives should contribute
optimising airspace usage while facilitating the Shared/ Reference
towards regional goals [3]. Capability of large-scale contingency
Business Trajectories from the Long-term Planning Phase to the
operations in case of prohibitive weather and airspace closures,
Pre-flight Execution Phase.
and scalability to accommodate future growth are deemed
AMAN, SMAN and DMAN fully integrated into CDM processes essential attributes, and the system should be robust and have a
toenhance the CDM-mediated business processes of airport high availability, while at the same time encompassing graceful
stakeholders by considering the emerging enablers offered by degradations when unavoidable. The system should be natively
AMAN, SMAN and DMAN systems (such as the user Target applicable globally to all stakeholders and capable of
Time of Arrival). institutionalising green procedures (such as those championed by
Asia and Pacific Initiative to Reduce Emissions (ASPIRE) today
Introduction of UDPP and Collaborative Departure Sequence by or those arising from the Asia-Pacific Economic Cooperation
automated application of the prioritisation rules provided by each (APEC) ATM Emissions Reduction initiative) without disruption
user. to non-participating traffic. ASPIRE has a pragmatic approach
Introduction of CDM in the SWIM Environment seeks to integrate that focuses on specific city-pairs in line with the objectives for
collaborative planning and CDM processes within the SWIM regional ATFM in Asia-Pacific.
network.
6.6.1. Asia-Pacific regional characterisation
A number of SESAR validation exercises have been conducted
[121]. The degree to which SESAR research aspirations have The Asia-Pacific region is characterised by a high proportion of
been achieved varies across workpackages as documented in the international traffic, necessitating regional collaboration to
SESAR annual reports [122]-[123]. achieve significant ATFM objectives. The lack of a single
regulatory authority implies a requirement for harmonisation of
6.5. Africa & Middle-East diverse, but equally valid, operational concepts and procedures
for efficient regional, cross-border operations. Additionally, the
To manage a projected traffic growth of 5.4% per annum, and the presence of smaller FIRs in South East Asia (characterised by
peak demand expected during the 2010 FIFA Soccer World Cup, transit times of less than 20 minutes) is also a peculiar aspect to
South Africa procured the Thales FLOWCAT ATFM system and be considered in the design of a successful regional ATFM
the Metron Harmony Slot Manager for its Central Airspace initiative, as it is the prevalence of islands or archipelagos where a
Management Unit in 2007 [80]. higher density of CNS systems is located, in contraposition to
FLOWCAT fuses data from several ATM systems into a “oceanic” environments, also very diffuse in the region. Within
consolidated national or regional view of current and forecast the region, an emerging peak saturation of enroute sectors is
traffic. This, and the use of the Eurocontrol CFMU ATFM increasingly recurring (e.g. 20-30% overload in some Bangkok
message set, underpin plans to expand ATFM operations to the ACC sectors), and the surveillance performance is inadequate at
Southern African region. Key features include the ability to [81], some busy metering fixes, resulting in a reliance on vertical
[82]: separation. In relation to the various national and largely non-
coordinated modernisation processes, the ATM technology
 Implement FUA concepts by modelling the online employed is considerably fragmented, posing challenges to the
reservation or release of airspace at short notice to match achievement of a region-wide, real-time common situational
fluctuating peak demand;  awareness, as well as to sustaining, up-scaling and automating the
communication infrastructure required for Collaborative Decision
 Display thunderstorm data and imagery;  Making (CDM). The region, nevertheless, is characterised by a
 Probe aircraft flight plans against moving thunderstorms and relatively young fleet, equipping state-of-the-art CNS+A systems
automatically propose re-routes around them, taking into supporting enhanced navigation accuracy, data-link
account aviation separation standards;  communications and dependent surveillance.
 Predict and smooth imbalances in controllers’ traffic-related
6.6.2. Regional ATFM Organisational Structure
workload using multi-flight “what-if” modelling features to
model delays and re-routes;  A political mandate for a centralised Regional Network Manager
(akin to those in Europe and the US) in Asia-Pac is unlikely to be
 Interface with a Slot Manager and enable CDM.  feasible in the short term; nor may this be the best way to proceed
operationally given the diversity of airspace in the region and the
The United Arab Emirates has implemented a web-based CDM fact that each individual authority remains best-placed to judge
Departure Flow Management program to cope with the spacing the operational and safety issues pertinent to its own airspace.
requirements of downstream FIRs since, as a hub, over 75% of Recent developments have seen the emergence of a more realistic
their movements is international traffic [83]. They are now alternate organisational structure that consists of a network of
exploring broader collaboration in the region. interconnected, co-operating ATFM nodes servicing sub-regions
but sharing CDM information in order to achieve common
6.6. Virtual ATFM Centre in the Asia-Pacific regional goals [57]. Observe, in particular, the sharing of
The implementation of a regional ATFM in the Asia-Pacific information amongst the sub-regional ATFM nodes in the Asia
context needs to be approached from the perspective of Seamless Pacific model. The ATFM nodes not only collate flight data for a
ATM. Indeed, the Asia Pacific Seamless ATM Plan promotes sub-region – they can also share it with other sub-regions in
“sub-regional ATFM based on system-wide CDM” [58]. Several
24 T. Kistan et al. / Progress in Aerospace Sciences

addition to distributing it within a sub-region (subject to business- rule filters such as removing military traffic).

Fig. 19. Contrasting Regional ATFM Structures in Europe (left) and Asia-Pacific (right).

Despite the contrast observed above, the involvement of several has not yet been reached on all points – the lack of common
parties in multiple projects to define a Regional ATFM separation criteria for RECAT-2 being an example [119]:11.
Operational Concept for Asia-Pacific will likely continue to
provide access to international best-practice to inform the 6.6.4. ATFM-related ASBU Elements
deliberations of the ICAO Asia/Pacific ATFM Steering Group.
For example, the European Union Association of Southeast Asian The four Performance Improvement Areas identified by ICAO are
Nations (ASEAN) Air Transport Integration Project (AATIP) [17], [18]:
includes the development of cross-border international ATFM  greener airports;
concepts, while the FAA is heavily involved in the APEC Air
Traffic Management Emissions Reduction Project, which has a  globally interoperable systems and data – through Globally
strong focus on ATFM and CDM as enablers for emissions Interoperable SWIM;
reduction.
 optimum capacity and flexible flights – through global
CDM technologies such as Web-Services, Video-Conferencing, collaborative ATM;
Collaborative Workbenches and traditional AFTN enable the
concept of a “virtual” ATFM node. Individual ANSPs contribute  efficient flight path – through TBO.
staff and systems that remain located in their native facilities but ATFM has developed progressively over the last three decades
which share information to establish common situational [19]. It has been inferred from the European modernisation
awareness, provide drill-down analysis and “what-if” modelling initiative that key steps are necessary to be able to predict traffic
capabilities and execute ATFM measures that impact traffic in loads for the successive days with good accuracy. It has also been
the sub-region. Whether we consider each local collection of identified that there is a need to move from conservative measures
physical resources to be a “local virtual ATFM node” – or the that define traffic as rate of entry into a given airspace (and not as
sum of all local collections of physical resources for a sub-region departure slots) to measures implemented before take-off that take
to collectively form the “virtual sub-regional ATFM node” (as into account the capacities in a wider area. The significance of
indicated below) is a matter of semantics – the structure and proposing alternative routings rather than only a delay diagnosis
functionality remain the same. has also been identified as a means of preventing over-
reservations of capacity. It has been identified as part of ASBU,
that in order to better regulate traffic flows ATFM needs the
following measures [19]:
a) departure slots ensuring that a flight will be able to pass the
sectors along its path without generating overflows;
b) rate of entry into a given piece of airspace for traffic along a
certain axis;
c) requested time at a way-point or an FIR/sector boundary
along the flight;
d) miles-in-trail figures to smooth flows along a certain traffic
axis;
e) re-routing of traffic to avoid saturated areas;
Fig. 20. A distributed, virtual ATFM node. f) sequencing of flights on the ground by applying departure
time intervals;
6.6.3. SESAR/NEXTGEN ICAO Harmonisation g) level capping;
The FAA and Eurocontrol are actively working towards achieving h) delaying of specific flights on the ground.
interoperability between the NextGen and SESAR programs
As part of block 0 upgrades, improved flow performance through
[119]. Harmonisation plans include co-ordination through ICAO
planning based on a network-wide view has been proposed.
standards and consideration of ASBU modules [119]. Agreement
25 T. Kistan et al. / Progress in Aerospace Sciences

Collaborative ATFM measures will be used to regulate peak Table 3. ICAO ATFM SG critical ATFM-related ASBU elements.
traffic flows involving departure slots. Further operational
ATFM-related ASBU Element Priority
measures identified are managed rate of entry into a given piece
of airspace for traffic along a certain axis, requested time at a B0-ASUR Initial Capability For Ground Surveillance 1
way-point or an FIR/sector boundary along the flight and use of
B0-RSEQ Improve Traffic Flow Through Runway
miles-in-trail to smooth flows along a certain flow. The 2
Sequencing (AMAN/DMAN)
techniques and procedures depend on the inferences derived from
the ATM modernisation programmes and state-of- the-art of the B0-TBO Improved Safety And Efficiency Through The
1
current ATFM systems in place in certain regions of the world. Initial Application Of Data Link En-Route
As part of block 1 upgrades, enhanced flow performance through B0-FICE Increased Interoperability, Efficiency And
1
network operational planning has been proposed. ATFM Capacity Through Ground-Ground Integration (AIDC)
techniques identified in this block upgrade include efficient
B0-CDO Improved Flexibility And Efficiency In Descent
management of airspace, regulated traffic flows including initial 2
Profiles (Continuous Descent Operations - CDO)
user driven prioritisation processes for collaboratively defining
ATFM solutions based on priorities (commercial or operational B0-CCO Improved Flexibility And Efficiency Departure
2
driven). Profiles - Continuous Climb Operations (CCO)
As part of block 2 upgrades, increased user involvement in the B0-APTA Optimization Of Approach Procedures Including
1
dynamic utilization of the network has been identified. The Vertical Guidance
detailed improvements include the introduction of CDM
B0-ACDM Improved Airport Operations Through Airport-
applications that will be supported by SWIM, permitting airspace 2
Collaborative Decision-Making (A-CDM)
users to manage competition and prioritisation of complex traffic
flow management solutions when the network, or its nodes, is not B0-SURF Safety and Efficiency Of Surface Operations
3
able to provide sufficient capacity for the required demand. (A-SMGCS)
Further improvements identified include: B0-NOPS Improved Flow Performance Through Planning
1
Based On A Network-Wide View
 taxi schedules will be integrated with ANSP AMAN and
DMAN capabilities, to improve the execution of overall Table 4. APANPRIG Critical ATFM-related ASBU Elements.
ATFM strategies;
ASBU Element ATFM Related Features
 flight deck operations will be enhanced by taxi route
guidance and adoption of synthetic vision displays; Flexible Use Airspace (FUA), User Preferred
B0-FRTO Enhanced Routes (UPR), Dynamic Airborne Re-route
 improved data sharing to enhance the available ATFM En-route Trajectories Planning (DARP), Collaborative Decision
information; Making (CDM)
B0-FICE Ground – ATS Inter-facility Data Communications
The identified enhancements will improve safety, efficiency and
Ground Integration & (AIDC) – in support of Dynamic and Flexible
DCB. Furthermore, these activities are intended to minimise taxi Interoperability ATS Route Management
times, within the context of any higher level ATFM/airport
resources strategy resulting in reduced fuel burn and B0-DATM Digital
System-Wide Information Management
Aeronautical
environmental impacts. (SWIM) – as an enabler for B0-NOPS
Information
Air Traffic Flow Management (ATFM).
The ICAO ATFM Steering Group (SG) has ranked the Interim Management (AIM)
ATFM-related ASBU Elements in order of priority as listed in “Inter-linked and networked ATFM nodes
B0-NOPS Network
Table 3. The Asia Pacific Air Navigation Planning and between ANSPs should be developed to serve
Flow Management
Implementation Regional Group (APANPRIG) has its own various sub-regions”
selection of critical and recommended ASBU upgrades which B0-RSEQ Runway Arrival Manager (AMAN), Departure
also reveals strong links to ATFM (Table 4). As ATM systems Sequencing Manager (DMAN).
move towards supporting these ASBU capabilities, their B0-ACDM Airport Airport Collaborative Decision-Making (A-
associated ATFM nodes should be planned with the CDM CDM).
corresponding capabilities to manage them.
Aerodrome warnings, including windshear.
B0-AMET
World Area Forecast Centre (WAFC),
Meteorological
Volcanic Ash Advisory Centre (VAAC), and
Forecasts, Warnings
Tropical Cyclone Advisory Centre (TCAC)
and Alerts
forecasts.
Extended arrival metering, integration of
B1-RSEQ Runway
surface management with departure
Sequencing
sequencing.
Integrated ATFM including airspace
B1-NOPS Network
management, user driven prioritisation and
Flow Management
collaborative ATFM solutions.
Synchronisation of traffic flows at merge
B1-TBO Trajectory points through controlled time of arrival
Based Operations capability and airport applications such as D-
TAXI).
B1-AMET
Meteorological Weather information supporting automated
Forecasts, Warnings decision support or aids.
and Alerts
26 T. Kistan et al. / Progress in Aerospace Sciences

moment a CNS system fault resulting in an unsafe

7. ATFM systems in the CNS+A context airport/airspace condition is detected.
ATM is undergoing an evolutionary transition, for which Based on the above definitions, two separate models for the time
progressively higher levels of information sharing and automation responses associated with the Prediction-Avoidance (PA) and
will support a broadening of the scope from tactical interventions Reaction-Correction (RC) functions performed by the integrity
towards more strategic initiatives. Since ATFM was specifically augmentation system can be defined. The PA time response is
conceived to address the strategic online timeframe ATFM given by:
systems have a central role in this transition, ultimately
establishing the strategic time goals that maximise the traffic flow ∆T ∆T ∆T ∆T 39
optimality. In order to be successful, this evolutionary transition
depends crucially on very high levels of predictability and where ∆T is the time required to predict a critical condition,
integrity. Although the level of predictability is growing thanks to ∆T is the time required to communicate the predicted
the increasing adoption of 4DT and to the more copious and CNS system failure to the ATFM network and ∆T is the
timely information exchanged with the ground and between time required to implement an action that prevents the failure
aircraft, a number of potential disruptions are still not adequately from compromising the safety of traffic flow. Consequently we
addressed as part of current ATFM system implementations. In must have ∆T TTC. If the available avoidance time
this perspective the missing element is represented by integrity ∆T is not sufficient to implement an adequate action (i.e.,
monitoring and augmentation technologies, featuring adequate ∆T TTC), the traffic involved in that particular
predictive and reactive mitigation measures. The system integrity airspace/airport will inevitably encroach on critical conditions
metrics of specific value for ATFM are represented by causing losses of situational awareness or separation and/or
communication, navigation and surveillance performance unacceptable accuracy degradations. In this case, the RC time
parameters. response applies:
The implementation of integrity monitoring and augmentation ∆T ∆T ∆T ∆T 40
strategies tailored for the ground-based component of the CNS+A
paradigm is supported by Performance-Based Communication, where ∆T is the time required to detect a CNS system
Navigation and Surveillance (PBC/PBN/PBS) concepts and failure, ∆T is the time required to communicate the
system design methodologies. In particular, CNS systems failure to the ATFM network and ∆T is the time required to
performances to be considered are accuracy, integrity, availability perform a corrective action. Further evolutions are possible by
and continuity required for the proposed operations in the context including an early correction action as soon as the condition
of a particular airspace. Ideally, communication systems shall ∆T ≤ TTC is violated. In this case, the direct Prediction-
meet specific performance standards in terms of transaction time, Correction (PC) time response is:
continuity, availability, message integrity, safety and functionality ∆T ∆T ∆T ∆T 41

for the intended operation in a particular airspace. Navigation
systems shall meet specific performance requirements in terms of The implementation of a model-predictive integrity monitoring
accuracy, continuity, availability, integrity and functionality for and augmentation process is dependent on the development of
the intended operation in a particular airspace. Surveillance CNS system performance and error models as depicted in Fig. 21.
systems shall meet specific performance standards in terms of
surveillance data delivery time, continuity, availability, integrity,
accuracy, safety and functionality needed for the intended
CNS+A System Performance Model CNS+A System Error Model
Sector i1 LOS  Sector i1 LOS 
operation in a particular airspace. RADAR Communication RADAR Communication
BLOS  BLOS 
Sector i2 Sector i2
Communication Communication
Integrity involves the ability of the CNS system to provide timely ADS‐B
Radio/Instrument/Satellite 
ADS‐B
Radio/Instrument/Satellite 
and usable warnings to users when the system may not be used Sector in
ADS‐B
Navigation Sector in
ADS‐B
Navigation

for the intended operation or phase of flight. To implement the RADAR

Non‐Cooperative and 
Cooperative Surveillance RADAR
Non‐Cooperative and 
Cooperative Surveillance

required levels of integrity of CNS systems, the following alert Performance Parameters Errors in Measurments

states are defined [124]-[126]: Integrity Indicators for Demand‐Capacity Balancing (4D)

 Caution Integrity Flag (CIF): a predictive annunciation Dynamic Adapatation of Airspace

that the CNS systems performance currently available to
support the intended ATM Level of Service (LoS) and
capacity in a particular region is going to exceed the Fig. 21. Role of CNS performance and error models in the
performance thresholds specified for the current and integrity monitoring and augmentation process.
planned operational tasks (alert status).
Two different models can be defined for determining the
 Warning Integrity Flag (WIF): a reactive annunciation performance of a CNS system and the errors affecting the
that the CNS systems performance currently available to measurements. For example, Primary Surveillance Radars (PSR)
support the intended ATM LoS and capacity has can be modelled using the maximum range at which the system
exceeded the performance thresholds specified for the can determine a target of a specified size, accuracy of target
current and planned operational tasks (fault status). location measurement in terms of range and bearing, recognise
In relation to the aforementioned definitions, the following Time- the type of target, etc. Hence the performance model is different
to-Alert (TTA) parameters can be introduced: for each system. The deviations between the performance and
error models are assessed and translated into time. Integrity is
 Time-to-Caution (TTC): the minimum time allowed for translated into time (timeframe to violation) and serves as an
the caution flag to be provided to the user before the indicator for dynamic adaptation of airspace based on CNS
onset of a CNS fault resulting in an unsafe systems performances. Capacity changes in both space and time
airport/airspace condition. are thus clearly dependent on performances of CNS systems.
 Time-to-Warning (TTW): the maximum time allowed
for a warning flag to be supplied to operator from the
27 T. Kistan et al. / Progress in Aerospace Sciences

Research is on-going, on a regional as well as global scale, for

7.1. Impact of Unmanned Aircraft Systems (UAS) demonstrating the feasibility of UTM concepts. In the USA,
In the CNS+A framework, the general case is that of multiple NASA, along with UAS operational and regulatory stakeholders,
manned/unmanned aircraft performing either cooperative or non- is researching prototype technologies for a UTM system that
cooperative surveillance. In terms of granting the required levels could develop airspace integration requirements for enabling safe
of operational safety when considering the integration of and efficient UAS operations (especially targeting low-altitude
unmanned traffic in airspace characterised with dense air traffic tasks performed by unmanned aircraft). The UTM system
and high levels of ATS, the emphasis is on CNS+A equipment provides to remote pilots information needed to maintain
that can meet strict performance requirements while also separation from other aircraft by reserving areas for specific
supporting enhanced ATM functionalities. These systems will routes with consideration of restricted airspace and adverse
enable the UAS Traffic Management (UTM) paradigm including weather conditions. The UTM concept is illustrated in Fig. 22
UAS Traffic Flow Management (UTFM) functions. In particular, highlighting CDM between UAS operation centres and UTM
in order to grant the required surveillance performance and stakeholders. Strategic separation coordination and some tactical
autonomous separation maintenance and collision avoidance separation management is considered as part of UTM in addition
capabilities for unrestricted access to controlled airspace, UAS to contingency management, surface operation management and
surveillance equipment involve a combination of non-cooperative capacity-demand balancing. In the long term, UTM research is
sensors, including active/passive Forward-Looking Sensors (FLS) envisaged to integrate, in particular, the functionalities for
and acoustic sensors, as well as cooperative systems, including trajectory prediction and negotiation, conflict detection and
ADS-B and Traffic Collision Avoidance System (TCAS). In resolution, separation/spacing monitoring and optimisation for
order to overcome the limitations in current Airborne Collision mixed manned/unmanned aircraft traffic both within individual
Avoidance Systems (ACAS), the FAA has funded research on controlled airspace sectors and across multiple sectors belonging
Next Generation Airborne Collision Avoidance System NextGen to a FIR, also considering the inbound/outbound flows from
- ACAS X. Some of the key limitations in current ACAS include adjacent FIRs. Additionally, UTM supports functions associated
inadequate/improper coordination with ATM/ATFM systems, with airspace design and management, geo-fencing, congestion
lack of terrain/ground and obstacle awareness, mostly only range- management, authenticated operations and weather predictions.
based presentation of traffic and unsuitability to include Remotely As a significant portion of the UAS fleet is specifically employed
Piloted Aircraft Systems (RPAS) in the framework. Therefore, on an opportunity-basis with little to no advance planning, one of
taking advantage of recent advances in computational techniques the key challenges that needs to be addressed is the access of this
and to keep pace with planned future operational concepts, ACAS opportunity UAS traffic to controlled airspaces and the
X is foreseen to replace TCAS in the long-term. subsequent implications for the design and development of future
Voice/data/network radio communication system and satellite ATFM systems.
communication systems data links act as the backbone for Line-
of-Sight (LOS) and Beyond LOS (BLOS) communications
between air-to-air, air-to-ground and ground-to-ground systems.

Fig. 22. UTM concept [136].

present, traditional standards and regulations are in place through

7.2. ATFM system certification ICAO Annexes 8 and Doc 9760 for the individual components of
As part of the progressive transition to an integrated and more the system, with more emphasis on the airborne systems than for
strategic CNS+A scenario contemporary certification the ground-based systems. These standards need to evolve to
considerations for CNS/ATM systems will become applicable to account for the functional integration of the complex integrated
ATFM systems as well. CNS+A systems. A gap is observed for standards that could
certify ground systems, with only the Radio Technical
In the CNS+A context, ATM and avionics systems are integrated Commission for Aeronaut (RTCA) DO-278A being available for
in various aspects with the airborne systems, with a functional software certification but only in a traditional CNS/ATM
integration between air, ground and space systems. Integrated perspective (i.e., not addressing the CNS+A context).
systems demand greater safety verification efforts, considering Furthermore, while RTCA DO-178C and RTCA DO-254 provide
the complex nature of the interactions taking place across these a globally accepted framework for safety assessment of avionics
systems. The verification, validation and certification of the systems, the requirements for ground CNS/ATM system
individual components as well as the “integrated” CNS+A
systems capable of TBO/PBO are crucial for future systems. At
28 T. Kistan et al. / Progress in Aerospace Sciences

certification are typically not mandated by national and (IBO). The regulatory framework and industry standards should
international authorities. also evolve as part of the comprehensive certification framework
of integrated CNS+A.
Certification requirements for the organisations engaged in
aircraft and component design, production, maintenance and The certification framework further requires a means of
maintenance/crew training are well established in the regulations compliance which describes how compliance will be
of the European Aviation Safety Agency (EASA), FAA demonstrated. The current EASA Part 21 acceptable means of
(Regulations Parts 21, 145 and 147) and other national aviation compliance provides the basic compliance codes for type
authorities, together with licensing requirements of pilots and certification. This should be further developed to capture
aircraft maintenance personnel. Continuing airworthiness integrated CNS+A requirements. An example of certifying
management requirements are written into the EASA Part M and ground-based ATM/ATFM systems for the CNS+A concept
FAA Part 43 regulations. Similar regulations should be would require the “integrated” airborne system interfacing and
progressively developed and integrated for ground ATM system communicating with the ground-based system via data-links for
design, certification, production and life cycle management when IBO/TBO to undergo the means of compliance for certification
their functional integration with avionics systems is considered. including engineering evaluation, selected test method or
Current regulatory requirements to certify the aerodrome service combination of lab test, ground test, flight test and simulation,
providers and ANSPs as organisations only partially fulfil the design inspection and equipment qualification as part of the
certification requirements of the ground navigational aids, certification program.
surveillance and ATM systems. The responsibility for the
The cost of CNS+A certification for future ATFM systems may
commissioning certification of ground systems is entrusted to the
result in a regional roadmap approach similar to the ASBU
ANSP organisation. In contrast, the civil aviation authority of the
roadmap. The requirement for certification may depend on the
State of Design undertakes the design certification of aircraft
sophistication (degree of automation, level of integration etc.) of
while the authority of the State of Registry is responsible for the
the ATFM system being considered.
issuance of the certificate of airworthiness of aircraft.
With regard to personnel licensing, flight crew, aircraft 8. Conclusions and Future Work
maintenance technicians/ engineers, air traffic controllers, flight
operation offices/dispatchers and aeronautical station operators The analysis performed indicates that ATFM research needs to
are required to obtain licenses issued in accordance with ICAO progress in a number of areas of high multidisciplinary nature.
Annex 1. While some States require ground navigational aids and In particular, the harmonisation of ATM and avionic systems
ATM system maintenance personnel to obtain a company responsible for different phases of flight is crucial to enable the
approval certificate, it is not equivalent to the licensing process of practical implementation of gate-to-gate concepts. This
aircraft maintenance technicians/ engineers. Furthermore, the harmonisation elicits the adoption of integrated AMAN, DMAN
licensing of design and maintenance personnel of ground systems and Airport-CDM systems [57] and can be successfully
is not mandated by ICAO Annex 1. accomplished at various levels, ranging from common
The current certification status of the key players of CNS+A configuration, data formats, operational data, compatible
indicates the airborne component of CNS+A as fully certified algorithms for trajectory computation, runway allocation, QNH-
while the ground segment is not certified to the level of the based altitude correction to sophisticated automation assisted
airborne segment. Similar to the aircraft initial and continuing capabilities (such as integrated “what-if” scenario modelling).
airworthiness certification requirements, the ground equipment A first practical step in this direction is represented by a
and personnel certification would be essential for the finalisation of open standards (such as FIXM [57]:49 and Web
airworthiness support function that it provides to the aircraft. Services) for the secure sharing of data amongst different systems
The systems specifications of aircraft systems for CNS+A is and amongst different regional authorities (SWIM).
currently based on regional requirements. Future harmonisation, Technologically, increases in the levels of automation in
interoperability and seamless operation require a global consensus ATM/ATFM DSS will allow a relatively small number of flow
to be provided for the equipment/system mandates through ICAO managers to oversee major traffic flows over vast geographic
annexes for each nation’s airspace, standardised through regions. In this respect, areas of current technological research
established industry standards. and development activities include the application of heuristic
Various approaches for the development of a unified framework agent oriented systems and the evolution of HMIs, for instance
supporting integrated CNS+A system certification are currently supporting adaptive/cognitive capabilities. Cultural differences in
considered, addressing the initial airworthiness/performance decision making are well-acknowledged [88]-[94], with a widely
certification, initial commissioning certification, and continuing studied aspect being the difference between individualism and
airworthiness, performance and operational certification. collectivism [89]:6. This dimension is particularly relevant in the
Performance certification requirements for both airborne and non- Asia-Pacific region and is expected to have a bearing on the HMI
airborne (ground and space based systems), together with design and its configurability for different end-users in the region.
certification of data-link systems would also be vital in addition to These are, however, deeper-seated issues than those related to
aircraft/avionics certification considering the highly automated, cultural preferences in conventional HMI design. Adaptive HMI
interconnected and interdependent nature of future CNS+A. are the ones that autonomously adapt display formats and
Currently available safety assessment and system development available actions to the current goals and abilities of the user by
processes which support aircraft certification are required to be monitoring user status, the system task, and the current situation
extended to the ground CNS system counterparts. A top-down [95]:50. They note, in particular, that adaptive HMI paradigms
systems engineering approach is recommended for the that are biologically inspired show potential to better correspond
development of the certification model for integrated CNS+A to human decision-making [95]:47, which has been characterised
encompassing adequate processes for safety assessment and as naturalistic and recognition-primed [96].
system development of integrated systems capturing both The integration of improved medium to short term traffic demand
airborne and ground-based components. The certification process forecasts that can account for seasonal variations, major events
is required to capture all functional components of future CNS+A, and local traffic and environmental factors is crucially important
including ATFM, PBO and 4D-TBO/Intent Based Operations and the goal in this perspective is the development of a model that
29 T. Kistan et al. / Progress in Aerospace Sciences

can extrapolate from IATA airport seasonal schedules to [12] "European ATM Master Plan ‐ The Roadmap for Sustainable Air
estimated runway slots on the day of operation. Machine learning Traffic Management", SESAR JU, Brussels, Belgium, 2012.
[13] R. Sabatini, A. Gardi, S. Ramasamy, T. Kistan, and M. Marino, "Novel
may play a role in improving existing heuristics. ATM and Avionic Systems for Environmentally Sustainable
Aviation", in proceedings: Practical Responses to Climate Change.
In parallel, an improved prediction of real-time traffic Engineers Australia Convention 2014 (PRCC 2014), Melbourne,
movements, particularly for the identification and tracking of Australia, 2014.
flows of aircraft must be pursued. This will ideally consist of a [14] R. Sabatini, A. Gardi, S. Ramasamy, T. Kistan, and M. Marino, "Next
straight-forward application of existing visual and predictive Generation Avionic and ATM Systems for Environmentally
analytic technologies to the ATFM domain. Sustainable Aviation", in proceedings: International Symposium on
Sustainable Aviation (ISSA 2015), Istanbul, Turkey, 2015.
Improvements in the modelling of the broader network effects of [15] R. Sabatini, A. Gardi, S. Ramasamy, T. Kistan, and M. Marino,
local tactical measures, starting with integrated AMAN/DMAN "Modern Avionics and ATM Systems for Green Operations", in
Encyclopedia of Aerospace Engineering, ch. eae1064, R. Blockley
operations between closely-spaced city pairs are also a necessary and W. Shyy, Eds., John Wiley & Sons, 2015.
step. In particular, relatively short international flight times in [16] "JPDO Trajectory‐Based Operations (TBO) Study Team Report",
various regions including the Asia Pacific mean that the AMAN NextGen Joint Planning and Development Office, 2011.
and DMAN planning horizons of a city pair straddling different [17] ICAO, "Background Information on Aviation System Block Upgrades
(ASBUs), Working Document on The Aviation System Block
countries could overlap. Upgrades, The Framework for Global Harmonization, Second
Novel technologies will accommodate User Preferred Routing or Version to The working Document for the Global Air Navigation
Industry Symposium (GANIS)", 2011.
Collaborative Trajectory Options implemented across [18] "ICAO Doc 9750 – Global Air Navigation Capacity & Efficiency Plan
international boundaries, as well as improved mechanisms for 2013‐2028", The International Civil Aviation Organization (ICAO),
regional CDM conducted in a multi-nodal ATFM network Montreal, Quebec, Canada, 2014.
[57]:50. As a consequence, regional ATFM centres needs to be [19] "Working Document for the Aviation System Block Upgrades, The
Framework for Global Harmonization", 2013.
timely notified of the real-time airspace capacity and constraints
[20] ICAO, "Annex 11 to the Convention on International Civil Aviation ‐
of all member countries [57]:50 and this will mandate a further Air Traffic Services", International Civil Aviation Organization
harmonisation of ATFM measures to long-range flights crossing (ICAO), Montreal, QC, Canada, 2001.
international boundaries, particularly where multiple measures are [21] ICAO, "Doc. 4444 ‐ Procedures for Air Navigation Services ‐ Air
applied to the same flight in different FIRs by different authorities Traffic Management", The International Civil Aviation Organization
(ICAO), Montreal, Canada, 2007.
[57]:49. [22] UK Aeronautical Information Services, "Wake Turbulence", NATS
The capacity-enhancing potential of DAM techniques needs to Aeronautical Information Circular P 3/2014, 2014.
[23] R. De Neufville, A. Odoni, P. Belobaba, and T. Reynolds, Airport
evolve beyond the temporary morphing of sector geometries. By Systems ‐ Planning, Design and Management, 2nd ed., McGraw‐Hill,
analogy to the dynamic separation standards foreseen for New York, USA, 2013.
RECAT-3, the separation standards applicable in enroute and [24] A. Marsden, "SESAR FACTSHEET ‐ Airport Operations
terminal airspace can be varied in space and time based on the Management", Eurocontrol Experimental Centre,, Bretigny‐sur‐
Orge, France, 2014.
CNS+A capabilities of the surveillance infrastructure and
[25] ICAO, "Doc. 9426 ‐ Air Traffic Services Planning Manual", The
avionics, yielding both baseline and opportunistic capacity International Civil Aviation Organization (ICAO), Montreal, Canada
increases. This relates to traffic count only and the impact on the [26] S. Muresean, "AMAN Information Extension to En‐route Sectors ‐
controller’s cognitive workload also needs to be considered in the Concept of Operations", Eurocontrol, Bruxelles, Belgium, 2009.
context of the adaptive HMI research recommended in this [27] D. Pascarella, F. Gargiulo, A. Errico, and E. Filippone, An analytical
approach for optimal resilience management in future ATM
section. systems, Studies in Computational Intelligence vol. 616, pp. 415‐
425, Springer Verlag, 2016.
References [28] T. Kistan, M. J. Philippe, L. S. Takalani, and S. Maphanga, "A practical,
route‐based approach to airspace capacity management", in
[1] ICAO, "Doc. 9971 ‐ Manual on Collaborative Air Traffic Flow proceedings: 28th International Congress of the Aeronautical
Management", The International Civil Aviation Organization Sciences (ICAS 2012), Brisbane, Australia, 2012.
(ICAO), 2nd ed., 2014. [29] M. Janic and V. Tosic, "Terminal Airspace Capacity Model",
[2] ICAO, "Doc. 9965 ‐ Manual on Flight and Flow Information for a Transportation Research Part A: General, vol. 16, pp. 253‐260,
Collaborative Environment", The International Civil Aviation 1982.
Organization (ICAO), disclosed draft ed., Montreal, Canada, 2012. [30] E. E. Velazco, "Air traffic management: high‐low traffic intensity
[3] ICAO, "Doc. 9854 ‐ Global Air Traffic Management Operational analysis", European Journal of Operational Research, vol. 80, pp.
Concept", The International Civil Aviation Organization (ICAO), 45‐58, 1995.
Montreal, Canada, 2005. [31] D. Bertsimas and S. S. Patterson, "The Air Traffic Flow Management
[4] "NextGen Implementation Plan", Federal Aviation Administration Problem with Enroute Capacities", Operations Research, vol. 46, pp.
(FAA), Washington DC, USA, 2013. 406‐422, 1998.
[5] C. Nickol, "Environmentally Responsible Aviation (ERA) Project", [32] A. Majumdar, W. Y. Ochieng, J. Bentham, and M. Richards, "En‐route
2011. sector capacity estimation methodologies: An international
[6] P. Ky and B. Miaillier, "SESAR: towards the new generation of air survey", Journal of Air Transport Management, vol. 11, pp. 375‐387,
traffic management systems in Europe", Journal of Air Traffic 2005.
Control, vol. 48, 2006. [33] A. M. Churchill, D. J. Lovell, and M. O. Ball, "Evaluating a new
[7] "Clean Sky 2 ‐ Joint Technical Programme", European Commission, formulation for large‐scale traffic flow management", in
Brussels, Belgium, 2014. proceedings: 8th USA/Europe Air Traffic Seminar (ATM'09), Napa,
[8] "Long‐term Vision for the Future Air Traffic Systems ‐ Changes to California, USA, 2009.
Intelligent Air Traffic Systems", Study Group for the Future Air [34] E. P. Gilbo and S. B. Smith, "New method for probabilistic traffic
Traffic Systems, Collaborative Actions for Renovation of Air Traffic demand predictions for en route sectors based on uncertain
Systems (CARATS), 2010. predictions of individual flight events", in proceedings: 9th
[9] F. Collier, "Overview of NASA’s Environmentally Responsible USA/Europe ATM R&D Seminar (ATM 2011), Berlin, Germany,
Aviation (ERA) Project", in proceedings: NASA Environmentally 2011.
Responsible Aviation Project Pre‐Proposal Meeting. Washington, [35] P. U. Lee, C. Brasil, J. R. Homola, and T. Prevot, "Benefits and
DC, 2010. feasibility of the flexible airspace management concept: a human‐
[10] J. Bronsvoort, G. McDonald, R. Potts, and E. Gutt, "Enhanced descent in‐the‐loop evaluation of roles, procedures, and tools", in
wind forecast for aircraft", in proceedings: 9th USA/Europe Air proceedings: 9th USA/Europe ATM R&D Seminar (ATM 2011),
Traffic Management Research and Development Seminar (ATM Berlin, Germany, 2011.
2011), Berlin, Germany, 2011. [36] D. Bertsimas, G. Lulli, and A. Odoni, The air traffic flow management
[11] Airservices Australia, "RNP Project Brisbane Green Stage One problem: An integer optimization approach, 13th International
Report", 2008. Conference on Integer Programming and Combinatorial
30 T. Kistan et al. / Progress in Aerospace Sciences

Optimization, IPCO 2008 vol. 5035 LNCS, pp. 34‐46, Bertinoro, [60] B. Favennec, E. Hoffman, A. Trzmiel, F. Vergne and K. Zeghal, “The
2008. Point Merge Arrival Flow Integration Technique: Towards More
[37] H. D. Sherali and J. M. Hill, "Configuration of airspace sectors for Complex Environments and Advanced Continuous Descent”, in
balancing air traffic controller workload", Annals of Operations proceedings:9th AIAA Aviation Technology, Integration and
Research, vol. 203, pp. 3‐31, 2011. Operations Conference (ATIO), Hilton Head, South Carolina, 21‐23
[38] J. Tang, S. Alam, C. Lokan, and H. A. Abbass, "A multi‐objective September 2009, AIAA 2009‐6921.
approach for Dynamic Airspace Sectorization using agent based [61] Eurocontrol, “Point Merge Integration of Arrival Flows Enabling
and geometric models", Transportation Research Part C: Emerging Extensive RNAV Application and Continuous Descent ‐ Operational
Technologies, vol. 21, pp. 89‐121, 2012. Services and Environment Definition”, version 2.0, 19/07/2010.
[39] W. Hall, A. Churchill, P. Lucic, and I. Hwang, "Dynamic terminal [62] T. Ishikawa, “Sharing of Experience from implementing the Japan
airspace configuration", in proceedings: 2013 Aviation Technology, ATFM and CDM meeting at ATM center” in proceedings: ICAO
Integration, and Operations Conference, Los Angeles, CA, 2013. ATFM Workshop, Beijing, October 2014.
[40] D. Delahaye, J.‐M. Alliot, M. Schoenauer, and J.‐L. Farges, "Genetic [63] CANSO, “Japan Civil Aviation Bureau (JCAB) selects NTT DATA for
algorithms for partitioning air space", in proceedings: 10th new ATFM/ASM system”, 18 September 2015,
Conference on Artificial Intelligence for Applications, Piscataway, https://www.canso.org/japan‐civil‐aviation‐bureau‐jcab‐selected‐
NJ, San Antonio, TX, USA, 1994, pp. 291‐297. ntt‐data‐new‐system
[41] P. Kopardekar, K. Bilimoria, and B. Sridhar, "Initial concepts for [64] CANSO, “DFS and Atech Deliver Air Traffic Flow Management
dynamic airspace configuration", in proceedings: 7th AIAA Aviation System for India”, 31 July 2014, https://www.canso.org/dfs‐and‐
Technology, Integration and Operations Conference (ATIO 2007), atech‐deliver‐air‐traffic‐flow‐management‐system‐india
Belfast, Northern Ireland, 2007. [65] ATMB/CAAC, “Synopsis of Air Traffic Flow Management In China”
[42] S. A. Martinez, G. B. Chatterji, D. Sun, and A. M. Bayen, "A weighted‐ in proceedings: ICAO ATFM Workshop, Beijing, October 2014.
graph approach for dynamic airspace configuration", in [66] G. Atkins, “New Zealand ATFM Status”, in Proceedings GLOBAL 5
proceedings: AIAA Guidance, Navigation, and Control Conference ATFM Conference, Brisbane, October 2009.
2007, Hilton Head, SC, 2007, pp. 1476‐1491. [67] W. Beeston, “ Australia ATFM Report”, in Proceedings GLOBAL 6
[43] D. Delahaye and S. Puechmorel, "3D airspace design by ATFM Conference, Phuket, February 2011.
Evolutionary computation", in proceedings: 27th IEEE/AIAA Digital [68] P. Martland, “Collaborative Decision Making Program ATFM Stage 1
Avionics Systems Conference (DASC 2008), St. Paul, MN, 2008, pp. – Update”, in Proceedings GLOBAL 7, Bali, January 2013.
3B61‐3B613. [69] R. Fahmi & P. Martland, “Joint Task Force
[44] N. L. Fulton and M. Westcott, "Simulating radio communication Indonesia‐Philippines‐Australia”, in Proceedings GLOBAL 7, Bali,
complexity in regional airspace design", in proceedings: 18th World January 2013.
IMACS/MODSIM Congress, Cairns, Australia, 2009, pp. 397‐403. [70] CAD, “ Hong Kong, China Status Update”, in Proceedings GLOBAL 6
[45] J. Zhang, C. Zhang, and M. Hu, "Airspace Behavior Modeling ATFM Conference, Phuket, February 2011.
Considering Flight Intent", in proceedings: 28th Digital Avionics [71] T. Piyawut, “ Thailand ATFM Status Update”, in Proceedings
Systems Conference: Modernization of Avionics and ATM‐ GLOBAL 6 ATFM Conference, Phuket, February 2011.
Perspectives from the Air and Ground (DASC 2009), Orlando, FL, [72] CANSO, “Bangkok – Singapore CDM Project”, in Proceedings
USA, 2009, pp. 7.A.21‐7.A.210. GLOBAL 7, Bali, January 2013.
[46] N. L. Fulton, M. Westcott, and S. Emery, "Influences of [73] ICAO, “General Situation of ATC Support Capability at South China
communication structural complexity on operational safety in Sea Area” in Minutes of the Second Meeting of the South China Sea
regional airspace design", Safety Science, vol. 49, pp. 1099‐1109, Major Traffic Flow Review Group”, Haikou, China, 22‐24 July 2015.
2011. [74] Eurocontrol, Air Traffic Flow and Capacity Management Strategy,
[47] J. Tang, S. Alam, C. Lokan, and H. A. Abbass, "A multi‐objective CFMU/URB/ATFCM‐STRAT‐01‐00, Edition 1.2, 01 April 2004.
evolutionary method for dynamic airspace re‐sectorization using [75] Wambsganss M.C., “Collaborative Decision Making in Air Traffic
sectors clipping and similarities", in proceedings: 2012 IEEE
Congress on Evolutionary Computation, CEC 2012, Brisbane, QLD, Management” in Transportation Analysis: New Concepts and
2012. Methods in Air Traffic Management, Bianca L., Dell’Olms P., Odoni
[48] D. Zhang, Y. Chen, H. Bi, and Z. Song, "Airspace sectorization based A.R. eds, 2001.
on weighted graph spectral bisection algorithm", in proceedings: [76] JPDO, Concept of Operations for the Next Generation Air
12th COTA International Conference of Transportation Transportation System, Version 2.0, 13 June 2007.
Professionals: Multimodal Transportation Systems ‐ Convenient,
Safe, Cost‐Effective, Efficient, CICTP 2012, Beijing, 2012, pp. 1924‐ [77] JPDO, Concept of Operations for the Next Generation Air
1935. Transportation System, Version 3.2, 13 June 2011.
[49] Y. Chen, H. Bi, D. Zhang, and Z. Song, "Dynamic airspace [78] Sang‐ngurn N., Richardson J., Choowong T. & Tantimekabutm P.,
sectorization via improved genetic algorithm", Journal of Modern “The Bay of Bengal Co‐operative Air Traffic Flow Management
Transportation, vol. 21, pp. 117‐124, 2013. System (BOBCAT)” in The Journal of Air Traffic Control, Volume 49,
[50] P. Flenel and J. Pearson, "Automatic airspace sectorisation: a No. 3, Fall 2007.
survey", Uppsala University Deliverable D2Y4, Contract 08‐ [79] ATNS, Contact ATNS Newsletter, June 2003.
1214447‐C, Uppsala, Sweden, 2013. [80] ATNS, CAMU Press Release, 2007.
[51] V. Sciandra, J. Wei, I. Hwang, and W. D. Hall, "Benefit analysis of a [81] S. Ratcliff, “South Africa ATFM Status”, in Proceedings GLOBAL 5
sector design algorithm for terminal dynamic airspace ATFM Conference, Brisbane, October 2009.liff, “South Africa ATFM
configuration", in proceedings: 2013 Aviation Technology, Status”, in Proceedings GLOBAL 5 ATFM Conference, Brisbane,
Integration, and Operations Conference, Los Angeles, CA, 2013. October 2009.
[52] J. Wei, V. Sciandra, I. Hwang, and W. D. Hall, "An integer [82] T. Kistan, “Advancing FLOWCAT through ATFM and CDM
programming based sector design algorithm for terminal dynamic Laboratories”, in Proceedings GLOBAL 5 ATFM Conference,
airspace configuration", in proceedings: 2013 Aviation Technology, Brisbane, October 2009.
Integration, and Operations Conference, Los Angeles, CA, 2013. [83] GCAA, “The need for innovative Flow Management. The UAE
[53] Y. Chen and D. Zhang, "Dynamic airspace configuration method Perspective”, in Proceedings Collaborative Global ATFM
based on a weighted graph model", Chinese Journal of Aeronautics, Conference, WAC, Madrid, March 2016.
vol. 27, pp. 903‐912, 2014. [84] SENEAM, “Development and Evolution of ATFM in Mexico”, in
[54] D. F. Zhang and Y. Z. Chen, "Airspace sectorisation via a weighted Proceedings GLOBAL 8 ATFM Conference, Cancun, November 2014.
graph model", Aeronautical Journal, vol. 118, pp. 267‐274, 2014. [85] ATECH, “SkyFlow ATFM System“, April 2014, accessed at
[55] Tang J, Alam S, Lokan C & Abbass HA, ” A Multi‐objective http://www.icao.int/SAM/Documents/2014‐
Evolutionary Method for Dynamic Airspace Re‐sectorization using SAMIG13/04%20Atech%20SkyFlow%20ATFM%20System.E.pdf
Sectors Clipping and Similarities”, WCCI 2012 IEEE World Congress [86] S. Ratcliffe, “Airbus ProSky Assisting ANSPs”, in Proceedings
on Computational Intelligence, 10‐15 June 2012, Brisbane. GLOBAL 8 ATFM Conference, Cancun, November 2014.
[56] Birdsall J, “Australian ATFM Status” in Proceedings GLOBAL 5 [87] Eurocontrol, “RECAT‐2 and RECAT‐3”, accessed at
ATFM Conference, Brisbane, 2009. http://www.eurocontrol.int/articles/pair‐wise‐separations‐pws‐
[57] ICAO Asia/Pacific ATFM Steering Group, “Asia Pacific Framework recat‐2
for Collaborative Air Traffic Flow Management”, Version 1.0 [88] Soonae P & Jung‐Soo Y, “Cross‐cultural differences in decision‐
September, 2015. making styles: A study of college students in five countries”, Social
[58] ICAO, Asia Pacific Seamless ATM Plan, Version 1.0, June 2013. Behavior and Personality, 2003,
[59] G. Boerger, P. Kappertz, P. Martin, C. Costello and B. Tveit, “AMAN http://findarticles.com/p/articles/mi_qa3852/is_200301/ai_n917
Operational Functions”, ORTHOGON AG EVP WP4, v1.1, Bremen, 4866/
Germany 2002.
31 T. Kistan et al. / Progress in Aerospace Sciences

[89] Güss CD, “Decision Making in individualistic and collectivistic [115] A. Gardi, R. Sabatini, and S. Ramasamy, "Multi‐objective
cultures”, accessed at optimisation of aircraft flight trajectories in the ATM and avionics
http://www.ac.wwu.edu/~culture/Guss2.htm context", Progress in Aerospace Sciences, vol. 83, 2016.
[90] Skovira RJ, “Cultural Influences and Corporate Decision Making: [116] R. K. Ahuja, T. L. Magnanti, and J. B. Orlin, Network Flows: Theory,
The Humanities/Information Systems Partnership”, 1993 accessed Algorithms, and Applications, new intl. ed., Pearson, 2014.
at [117] N. Y. Chen and B. Sridhar, "Management‐action‐embedded sector‐
http://www.eric.ed.gov/ERICDocs/data/ericdocs2sql/content_stor demand prediction models", Journal of Guidance, Control, and
age_01/0000019b/80/15/19/c1.pdf Dynamics, vol. 33, pp. 1892‐1898, 2010.
[91] Strohschneider S, “Cultural Factors in complex Decision Making”, [118] H. Zhang, Y. Xu, Y. Lei, and H. Liu, "Macroscopic Model and
accessed at Simulation Analysis of Air Traffic Flow in Airport Terminal Area",
http://www.ac.wwu.edu/~culture/Strohschneider1.htm Discrete Dynamics in Nature and Society, vol. 2014, 2014.
[92] Guo‐Ming C, “Seniority and Superiority: A Case Analysis of Chinese [119] NEXTGEN & SESAR, State of Harmonisation Document, December
Decision Making”, 1997, accessed at 2014.
http://www.eric.ed.gov/ERICWebPortal/custom/portlets/recordD [120] SJU, “Work Packages Summary” accessed at
etails/detailmini.jsp?_nfpb=true&_&ERICExtSearch_SearchValue_0 http://www.sesarju.eu/discover‐sesar/workpackages‐summary
=ED413619&ERICExtSearch_SearchType_0=no&accno=ED413619 on 15 April 2016.
[93] Mazadi Z, Ghasem‐Aghaee N & Nematbakhsh MA, “Agent’s Decision [121] SJU, “Validating SESAR”, accessed at
Making Based on Cultural Values”, 5th International Conference on http://www.sesarju.eu/innovation‐solution/validating‐sesar on 15
Electrical and Computer Engineering, ICECE 2008, 20‐22 December April 2016.
2008, Dhaka, Bangladesh. [122] SJU, Annual Activity Report, 31 March 2014, accessed at
[94] LeFebvre R & Franke V, “Culture Matters: Individualism vs. http://www.sesarju.eu/newsroom/brochures‐
Collectivism in Conflict Decision‐Making”, Societies 2013, 3, 128– publications/annual‐activity‐report‐2013 on 15 April 2016.
146; doi:10.3390/soc3010128. [123] SJU, 2014 Annual Report, ISBN 978‐92‐9216‐029‐6 – ISSN 2443‐
[95] Rothrock L, Koubek R, Fuch F, Haas M & Salvendy G, “Review and 5783 – doi:10.2829/11145, Luxembourg: Publications Office of the
reappraisal of adaptive interfaces: towards biologically inspired European Union, 2015.
paradigms” in Theoretical Issues in Ergonomic Science, Vol 3, No 1, [124] R. Sabatini, T. Moore, and C. Hill, "A new avionics‐based GNSS
pp 47‐84, Taylor and Francis, 2002. integrity augmentation system: Part 1 ‐ Fundamentals", Journal of
[96] Klein G., “Naturalistic Decision Making” in Human Factors, Vol. 50, Navigation, vol. 66, pp. 363‐384, 2013. DOI:
No. 3, June 2008. 10.1017/S0373463313000027
[97] J. C. Jones, D. J. Lovell, and M. O. Ball, "Combining Control by CTA [125] R. Sabatini, T. Moore, and C. Hill, "A new avionics‐based GNSS
and Dynamic Enroute Speed Adjustment to Improve Ground Delay integrity augmentation system: Part 2 ‐ Integrity flags", Journal of
Program Performance", 11th USA/Europe ATM R&D Seminar (ATM Navigation, vol. 66, pp. 501‐522, 2013. DOI:
2015), 2015. 10.1017/S0373463313000143
[98] J. D. Welch, J. Y. N. Cho, N. K. Underhill, and R. A. DeLaura, "Sector [126] R. Sabatini, T. Moore, and C. Hill, "Avionics‐Based GNSS
workload model for benefits analysis and convective weather Augmentation for Unmanned Aerial Systems Sense‐and‐Avoid",
capacity prediction", 10th USA/Europe ATM R&D Seminar (ATM 26th International Technical Meeting of the Satellite Division of the
2013), 2013. Institute of Navigation (ION GNSS+ 2014), Tampa, FL, USA, 2014
[99] J. D. Welch, J. W. Andrews, B. D. Martin, and B. Sridhar, [127] D.B. Work and A.M. Bayen, “Convex Formulations of Air Traffic
"Macroscopic workload model for estimating en route sector Flow Optimization Problems”, in proceedings: IEEE, Vol. 96, No. 12,
capacity", 7th USA/Europe ATM Research and Development December 2008.
Seminar (ATM2007), Barcelona, Spain, 2007. [128] P. Menon, G. Sweriduk, and K. Bilimoria, “New approach for
[100] H. Erzberger, "Transforming the NAS: The Next Generation Air modeling, analysis, and control of air traffic flow”, in AIAA J.
Traffic Control System", NASA Ames Research Center NASA/TP‐ Guidance, Contr., Dyn., vol. 27, no. 5, pp. 737–744, 2004.
2004‐212801, Moffett Field, CA, USA, 2004. [129] A.M. Bayen, R.L. Raffard & C.J. Tomlin, “Eulerian Network Model of
[101] "Modernising the European Sky", SESAR Consortium, Brussels, Air Traffic Flow in Congested Areas”, in proceedings: 2004
Belgium, 2011. American Control Conference, Boston (MA), USA, 2004.
[102] M. Baldwin, "Implementing SESAR, the Single European Sky's [130] A. M. Bayen, P. Grieder, G. Meyer, and C. J. Tomlin, "Lagrangian
Technological Pillar", ATC Global 2012, Amsterdam, Netherlands, delay predictive model for sector‐based air traffic flow", Journal of
2012. Guidance, Control, and Dynamics, vol. 28, pp. 1015‐1026, 2005.
[103] P. Ky and B. Miaillier, "SESAR: towards the new generation of air [131] B. Sridhar, T. Soni, K. Sheth, and G. B. Chatterji, "Aggregate flow
traffic management systems in Europe", Journal of Air Traffic model for air‐traffic management", Journal of Guidance, Control,
Control, vol. 48, 2006. and Dynamics, vol. 29, pp. 992‐997, 2006.
[104] "Concepts and Systems Integration Branch", NextGen Integration [132] P. K. Menon, G. D. Sweriduk, T. Lam, G. M. Diaz, and K. D. Bilimoria,
and Evaluation Capability, Multi‐Year Program Plan ANG‐E14, "Computer‐aided eulerian air traffic flow modeling and predictive
2012. control", Journal of Guidance, Control, and Dynamics, vol. 29, pp.
[105] "NextGen Implementation Plan", Federal Aviation Administration 12‐19, 2006.
(FAA), Washington DC, USA, 2013. [133] A. M. Bayen, R. L. Raffard, and C. J. Tomlin, "Adjoint‐based control of
[106] A. A. S. Galindo and K. Candelario, "Next Generation Air a new Eulerian network model of air traffic flow", IEEE
Transportation System (NextGen)", 2011. Transactions on Control Systems Technology, vol. 14, pp. 804‐818,
[107] UK Aeronautical Information Services, "Wake Turbulence", NATS 2006.
Aeronautical Information Circular P 3/2014, 2014. [134] D. Sun and A. M. Bayen, "Multicommodity Eulerian‐Lagrangian
[108] F. Guercio, "Air Traffic Control Congested Flow Optimisation large‐capacity cell transmission model for en route traffic", Journal
Models", PhD Thesis, University of Rome "La Sapienza", Rome, Italy, of Guidance, Control, and Dynamics, vol. 31, pp. 616‐628, 2008.
2000. [135] D. Sun, I. S. Strub, and A. M. Bayen, "Comparison of the performance
[109] C. Bosson, M. Xue, and S. J. Zelinski, "Optimizing Integrated Arrival, of four eulerian network flow models for strategic air traffic
Departure and Surface Operations Under Uncertainty", 11th management", Networks and Heterogeneous Media, vol. 2, pp. 569‐
USA/Europe ATM R&D Seminar (ATM 2015), 2015. 595, 2007.
[110] D. Bertsimas and M. Frankovich, "Air Traffic Flow Management at [136] S. Ramasamy, R. Sabatini, and A. Gardi, "A unified approach to
Airports: a Unified Optimization Approach", 10th USA/Europe ATM separation assurance and collision avoidance for flight
R&D Seminar (ATM 2013), 2013. management systems", 2016 IEEE/AIAA 35th Digital Avionics
[111] D. Bertsimas, M. Frankovich, and A. Odoni, "Optimal selection of Systems Conference, DASC2016, Sacramento (CA), USA, 2016.
airport runway configurations", Operations Research, vol. 59, pp.
1407‐1419, 2011.
[112] J. Avery and H. Balakrishnan, "Predicting airport runway
configuration", 11th USA/Europe ATM R&D Seminar (ATM 2015),
2015
[113] S. H. Kim, E. Feron, J. P. Clarke, A. Marzuoli, and D. Delahaye,
"Airport gate scheduling for passengers, aircraft, and operations",
10th USA/Europe Air Traffic Management Research and
Development Seminar, ATM 2013, 2013
[114] H. Belouadah, M. E. Posner, and C. N. Potts, "Scheduling with
Release Dates on a Single‐Machine to Minimize Total Weighted
Completion‐Time", Discrete Applied Mathematics, vol. 36, pp. 213‐
231, May 28 1992.

View publication stats