Collaborative ATFM Interface for Airlines (CAIA)

Collaborative ATFM interface for airlines (CAIA)

Gérard MAVOIAN, EUROCONTROL Experimental Center , Bretigny sur Orge, France

1

Abstract

In recent years, tactical Air Traffic Flow Management (ATFM) in Europe has moved towards a highly dynamic and adaptive process. Yet, it is commonly agreed by the main ATFM actors embracing the Central Flow Management Unit (CFMU), aircraft operators, airports and ATC providers that collaborative procedures need to be improved and fostered [1].

The CAIA project focuses on the collaboration between aircraft operators and the CFMU.

As a follow up to the collection of airline requirements, a decision support tool has been designed and prototyped based on the Collaborative Decision Making (CDM) common situation awareness concept ([2] and [3]). It aims at supplying airline users with enhanced real-time visibility on the ATFM situation within a time horizon of a few hours. Basically, the tool can be viewed as a kind of 4D congestion map. Several types of visualization have been defined. They encompass global views as well as 4D ATFM projections relative to specific flights combined with “what-if” re-routing facilities.

The principal objective of the tool is to increase ATFM efficiency by improving the co-ordination between two main planning activities taking place in the last few hours before flights departures:

• Route planning handled by airlines operators.
• Flow management monitored by the CFMU in collaboration with ATC providers.

Benefits expected from such a harmonization are potentially high. On the other hand, increasing airlines’ visibility on the ATFM process entails operational risks that have to be assessed before any implementation. Moreover, to demonstrate the operational feasibility of the concepts many issues have to be addressed. The most critical one is probably to provide a relevant decision making aid in the context of a highly dynamic and thus unstable environment.

Designing an ergonomic HMI is another issue. Classical 4D representation is combined with the high complexity of European ATFM environment and associated models.

An iterative process is necessary to refine and validate the tool and the underlying concepts. A simulation platform has been developed for this purpose. It allows faithful reproduction of ATFM events and environment.

Modeling flight plan management activities undertaken by airline operators using the CAIA tool is part of the validation plan. This is an initial step to further assess the impact on ATFM performances.

The final phase of the project will consist in carrying out live trials in co-ordination with stakeholders. British Airways is the candidate airline for this phase.

Introduction

Within the ECAC zone, flight plans have to be submitted to the CFMU for validation. Generally, depending on airline strategy and flight characteristics, the submission time lead relative to the departure time can vary from several months to a few hours

Collaborative procedures have been implemented between CFMU, aircraft operators and Flow Management Positions (FMP) to refine the ATFM plan during the pre-tactical phase (the two days before operations) based on both historical data and ATC capacity estimations for the next day. Subsequently, the CFMU notifies airlines about re-routing measures and temporary modifications of the routing scheme Although these procedures have proven their effectiveness their impact is limited to few identified traffic patterns due to the high uncertainty at this planning horizon.

It is during the day of operations that the dynamic “game” takes place. The objective is to optimize flight plans within a time window of a few hours before departure, taking into account real-time factors and constraints such as weather conditions or traffic congestion. The players are classical ATFM actors, the CFMU, aircraft operators, Flow Manager Positions (FMPs) and airports. They have specific objectives, business processes, strategies and awareness of the global situation. And they have to commonly determine for each flight:

• The best flight route including both horizontal and vertical profiles.
• The best take-off time.

Two decision making processes run in parallel to address these questions:

• Flight plan management handled principally by airlines
• Flow management monitored by the CFMU. Departure slots are allocated to part of the traffic to prevent from overloads in congested areas.

The figure 1 presents an overview of the decision making process (limited to the CFMU and airline actors):

Figure 1: flight planning and flow management

Flight route planning is carried out by airline operators supported usually by enhanced decision aid tools, including management of route catalogue, accurate 4D profile calculations and powerful route cost estimations. Many parameters are integrated such as wind, fuel consumption and en-route fees. Route optimization is regarded as an important business process and strategies vary from airline to airline. The main issue is to integrate the ATFM factor in a relevant way.

Once the route (including the flight level) has been defined, the flight plan is submitted to the CFMU for validation and distribution. The Initial Flight Planning System (IFPS) checks flight plans validity primarily regarding the routing scheme.

Then, flight plan is transferred to the CFMU system in charge of managing the real-time slot allocation process (ETFMS). Departure slots are sent to aircraft operators two hours before off- block times and are subject to modifications till approximately 30 minutes before take-off.

Obviously, humans stay in the loop and various procedures and protocols have been implemented to preserve the flexibility to treat specific cases. Airline flow managers are operators in charge of dialoging with the other ATFM actors in order to ensure slots compliance and possibly to reduce ATFM delays.

The black arrow in figure 1 closes the loop and symbolizes real-time adaptation of the traffic demand carried out by airlines in order to minimize ATFM delays. ATFM delay reduction must more than compensate possible extra-cost induced by route extension. During peak hours, up to 10 % of the traffic may be concerned.

The current process as described above presents important weaknesses.

The main one is that airlines primarily make the link between the two decision making processes. Due to the low visibility they have on the flow management activity, their efficiency to exploit spare capacity is far from being optimal. A high percentage of flight plan route modification (re-routings) undertaken to avoid congested zones and thus to reduce delays are proven to be unsuccessful. Moreover, competitive behaviors act as additional brakes.

The figure 2 illustrates this lack of efficiency. Distribution of ATFM delay gained per re-routing trial is shown below. The rate of failure is around 30 %, assuming (simplification) that a 10 minutes delay reduction is the threshold for successful action. These statistics have been produced by the CAIA project team from ATFM operational records. All route modifications applied to flights with significant initial ATFM delay have been considered.

Figure 2: re-routings delay reduction histogram

Thus, there is a clear issue raised by all stakeholders to rationalize the overall decision making process by realizing a real synergy between flow management and route planning processes. The CFMU has tried to implement some procedures and facilities for this purpose, but the impact is still limited.

Various orientations can be envisaged regarding in particular the distribution of the roles and responsibilities allocated to the different ATFM actors.

Whatever the strategy retained, a prerequisite is to involve the airlines much more in the process and thus improve their real-time awareness of the ATFM situation. Consequently, adapted tools have to be defined, validated and implemented. This is the purpose of the CAIA project.

The CAIA tool functionality

Based on airlines requirements, the objective of the CAIA project is to design and validate a decision support tool enabling aircraft operators to better integrate the ATFM factor in their route planning activity.

The CAIA tool can be viewed as a passive HMI merging in real-time information from two different sources: CFMU and airline systems. The figure 3 presents an overview of the tool functionality.

Figure 3: overview of CAIA tool

Three main kinds of dynamic ATFM visualizations have been designed and prototyped. They are detailed in the following subsections.

Real-time visualization of regulations.

A regulation is an explicit capacity restriction applied on a traffic volume during a time period. Traffic volumes can be can either basic ATC units such as en-route sectors or as more complex entities such as traffic flows. They can be activated, modified or cancelled in real-time by CFMU operators usually on FMP requests. Figure 4 presents a snapshot of the visualization:

Figure 4: real-time visualization of the regulations.

In the map, black and white (yellow and red in color) areas and associated labels represent graphically regulations known at a given time. The display is updated either periodically or on user requests. Color codes allow highlighting of recent changes in the ATFM panorama. Several kind of filters, such as the time horizon or levels of interest, are implemented to configure the display.

Display of regulations on airlines route options.

This function displays in real time all regulations which are likely to affect a flight along its possible 4D profiles considering different alternative routes. In addition to time period and capacity, maximum delay generated per regulation is supplied. A user can thus easily estimate ATFM delay on each route option. Both horizontal and vertical representations have been prototyped. Figure 5 presents an example of visualization of regulations on three different route options.

Figure 5: display of regulations on airline routes

The function can be triggered from an external application through an interface. This allows the route selection process to be carried out with support of dedicated airline flight plan management systems.

Visualization of congestion status on airline route options

Congestion is a more general notion than regulation. At an instant, only a limited number of regulations are active within the ECAC area (100 is the maximum). Congestion, on the other hand, concerns all entities such as sectors, airports and predefined traffic flows (traffic volumes) for which a capacity is declared. Congestion status for an entity is calculated basically by comparing estimated traffic load to capacity for a time period. Visualization of congestion enables users to better anticipate modifications of ATFM situations in the short term. It allows airline users to identify:

• ATC units which are likely to be regulated in the near future
• Areas with remaining spare capacity.

Visualization of congestion raises several issues:

• A flight is potentially concerned by many ATC units and traffic volumes. Visualization of all associated congestion status is not relevant. Consequently, intelligent filtering mechanisms have to be implemented based on both real-time information and analysis of historical data.
• Traffic loads and consequently congestion status are highly unstable measures. The display refreshment policy has to be defined carefully.
• Basically, congestion status is represented by a number of flights. What can be provided to users is either the raw information or derived data such as probabilities of regulation activation. Different options will be studied.

The function can be triggered from an external application as for the visualization of regulations on routes.

Validation strategy

Due to the high complexity of both operational and technical environments, a classical linear process is not well adapted for the implementation of such a tool. Many issues have to be addressed first such as:

• Providing a tool that increases airlines visibility on the ATFM process induces operational risks. Effects on operational working methods have to be evaluated to assess the impact on the overall ATFM process. Factors such as competitive behaviors have to be taken into consideration.
• Relevant decision making aids have to be provided to end users in the context of a highly dynamic process.
• Designing ergonomic HMI is quite ahard task. Classical 4D representation issues are combined with high complexity of ATFM environment models.

Consequently, particular attention has been paid to the definition of the validation strategy based on an incremental and iterative approach. Additionally, a validation platform has been developed to provide support for the activity.

The following diagram presents an overview of the process.

Figure 6: validation strategy

The following subsections detail some of the validation steps.

Simulations with end users

Reliability and stability of the information displayed strongly influence operational use of the tool.

The demonstration platform is used to reproduce a realistic decision making environment. A larger number of simulation sessions involving airlines operators are planned.

Scenarios are extracted from operational ATFM records.

Airline behavior modeling

As a follow up to the simulations, operational scenarios of use of the tool will be built. Airline behavioral models will be produced describing 3D route selection mechanisms including decisions time lead.

Obviously, these models have to integrate many factors such as the real-time information available, reliability and stability of the predictions, airlines global strategies and flight characteristics. Competitive behaviors should also be taken into consideration.

Implementation of models

For this purpose, a module is to be developed and connected to the validation platform. It implements the models built in the previous validation step and generates dynamically flight plan amendments simulating airline behaviors. Simulations will be processed for one, several and all airlines.

Operational assessment

To measure operational impact of the use of the CAIA tool, a new series of simulations will be planned. No operator will be involved as simulation of airline behavior will be automated. Full assessment implies studying the impact of the use of the CAIA tool on the following indicators:

• ATFM delays.
• Number of flight plan events generated.
• Time lead of flight plan events.
• Congestion status in ATC units.

Live trials

In the hypothesis that the operational assessment phase has enabled validation of CAIA tool concepts, live trials will be carried out to demonstrate, at least partially, technical feasibility. The CAIA platform will be connected to CFMU operational systems in shadow mode. British Airways is a candidate airline for being involved in the trials.

The validation platform

The figure 7 presents the general configuration of the simulation platform.

Figure 7: validation platform configuration

The ATFM simulator retrieves events relative to traffic and regulations from operational records. It provides to client applications dynamic ATFM data and events such as regulations, congestion status (calculated internally) and ATFM slots through several servers.

The airline system emulator allows users to roughly simulate facilities offered by the flight plan management systems to airline operators.

The platform has been built based on a distributed client/server architecture. It has been designed according to the following guidelines:

• Realism – An event-driven approach was retained to reproduce faithfully the ATFM dynamic adaptive process undertaken by the stakeholders.
• Efficiency - To cover the whole ECAC area, a large number of flights and events are manipulated. Performance issues have been addressed at early stage of the project.
• Flexibility – The platform offers light simulation facilities easily adaptable on client sites and requiring limited hardware and network resources. This allows flexible development and validation process carried out in close co-operation with end-users.

Technical choices have been made according to the above guidelines:

• C++ for the developments of the simulator.
• CORBA for the middleware.
• Concerning HMI developments, JAVA and LUCIAD Map have been used. LUCIAD Map is a product offering powerful Geographical Information System (GIS) JAVA software components.
• The platform runs on a PC with a Windows operating system.

Conclusion

The primary objective of the CAIA project is to address operational issues. Therefore, developing an the target operational tool is not the main goal. It is more to demonstrate the operational feasibility of enhancing airlines’ awareness of the ATFM process and the validity of the underlying CDM concepts. Enlarging the scope of the tactical ATFM activity to better integrate route management is the goalstake.

According to the potential benefits and the identified operational risks, an elaborated validation process has been designed that should allow the primary goal of the project to be achieved. Moreover, the validation infrastructure, including the simulation platform, is flexible enough to be reused by other projects dealing with tactical ATFM.

Abbreviations

ATCAir Traffic control

ATFMAir Traffic Flow Management

CAIACollaborative ATFM Interface for Airlines

CDMCollaborative Decision Making

CFMUCentral Flow Management Unit

ECACEuropean Civil Aviation Conference

ETFMSEnhanced Tactical Flow Management System

FMPFlow Management Position

GISGeographical Information System

HMIHuman Machine Interface

IFPSInitial Flight Planning System

References

[1] September 2000, Independent Study for the Improvement of ATFM, Contact: Philippe Jaquard (IGACEM France).

[2] P Martin, A Hudgell, S Vial, N Bouge, N Dubois, H de Jonge, & O.Delain, Potential Applications of Collaborative Decision Making, July 1999, EEC Note 09/1999

[3] Jasselin, Sureda-Perez et al., March 2001, A-CDM-D Final report, TEN 45601.

1