Requirements for an Allocation Center for the Flight Centric Air Traffic Control Concept ^†

Abstract

Nowadays, the allocation of aircraft to controllers is easily conducted. All aircraft entering a specific sector are automatically allocated to the responsible controller team. With increasing air traffic, it has become evident that this longstanding sector-based system needs to be replaced by innovative concepts such as Flight Centric ATC to meet the rapidly growing capacity demands. Flight Centric ATC requires novel systems for the allocation of aircraft to controllers, known as the Allocation Center. This Allocation Center has only been examined rudimentarily in current studies on Flight Centric ATC. The necessary requirements have not yet been taken into consideration in detail. This paper therefore both analyzes the general structure of an Allocation Center based on five high-level needs and also defines eleven operational requirements for such a center.

1. Introduction

With the suspension of travel restrictions imposed by nearly all countries in 2020 due to the COVID-19 pandemic, which significantly impacted the aviation sector, global aviation is now on the upswing and expected to return to the pre-crisis levels of 2019 soon. Initial forecasts for European air travel anticipate that, in all likelihood, this pre-crisis level will be attained in the 44 European Civil Aviation Conference (ECAC) states by the year 2025 [1]. However, this also means that as early as the coming year, the capacity issues that existed in Europe before the COVID-19 pandemic will become acute once again. Additionally, the Greener Sky initiative is becoming increasingly important, posing a challenge for European aviation to both align airspace capacity with demand and make air travel as environmentally friendly as possible [2]. Consequently, a renunciation of the traditional approach to air traffic control, where a controller team, consisting of an Executive and a Planner Controller, monitors an assigned sector, is necessary in the future. Under the Single European Sky ATM Research (SESAR) Programme, various concepts and enhanced air traffic controller (ATCO) tools are being researched to address these challenges [3]. One of these concepts is the Flight Centric ATC concept, initially published in [4] in 2001 and subsequently further developed, particularly by the German Aerospace Center, before becoming part of the SESAR2020/SESAR3 program in 2017 [5,6,7]. This concept involves an approach where the traditional sector boundaries, as they currently exist in Europe, are dissolved. Instead, a large, unified airspace is considered in which not only one controller team is responsible for air traffic control, but multiple controllers or controller teams share this task [8]. A key advantage of this approach is that aircraft can be distributed more evenly in terms of their workload across all Air Traffic Control Officers (ATCOs). Thus, the uneven workload, often observed between different sectors today, will be eliminated [9]. Various controller team compositions are possible. Similar to the traditional controller team, which consists of an Executive Controller and a Planner Controller, the Flight Centric ATC concept also enables a team consisting of an FCA Executive and an FCA Planner Controller to control aircraft [10]. The respective tasks of the two team members differ slightly from those in traditional ATC due to the new concept in place. Still, the Executive Controller is responsible for the safe and expeditious traffic flow of the allocated aircraft, while the Planner Controller is responsible for the planning and coordination of all incoming and outbound aircraft, as well as those within the Flight Centric ATC airspace [7]. It is also possible to slightly modify the conventional team composition by allocating one FCA Planning Controller to several FCA Executive Controllers or to specific traffic flows [7]. Another option is the Single-Person Operator. Here, a controller takes over the tasks of both the Executive and the Planner Controller. This is possible as the FCA concept changes the structure of the airspace to such an extent that the main task of the Planner Controller, the coordination of incoming and outgoing aircraft, is significantly reduced compared to the sectorized concept, which in turn leads to a significant reduction in the Planner’s taskload [7]. Various new tools and systems are needed to implement the Flight Centric ATC concept. On the one hand, these consist of support tools for the controllers to assist them in their task of air traffic control. These include, for example, a Less Impacted Flight Algorithm (LIFA), which in the event of a conflict between two or more aircraft identifies the aircraft that can solve the conflict with the least modification to the current flight trajectory. This is necessary as in an airspace in which several ATCOs are simultaneously responsible for controlling the aircraft, it must be clear at any time which of the controllers is responsible for conflict resolution in the event of a conflict [11]. Furthermore, the use of a Conflict Detection and Resolution (CD&R) tool, which not only points out potential conflicts to the controller but also creates one or more solution advisory for the controller with conflict resolution responsibility, is particularly useful in high- and very-high-complexity airspace [11]. On the other hand, new systems are needed to ensure that safe air traffic control is possible at any time. For this purpose, the Allocation Center is responsible, ensuring that each aircraft in the Flight Centric ATC airspace is allocated to a specific ATCO [12].

2. Current Status of Allocation Research

Researchers have been exploring the concept of FCA, focusing on how an Allocation Center and allocation strategies can be implemented. Preliminary approaches were defined, including ideas for using the Allocation Center for activities beyond specific aircraft allocation. The SESAR 2020 project PJ.10-W2-73 FCA introduced three levels of automation for allocation: automatic, semi-manual, and manual. Automatic allocation is carried out entirely by an Allocation Center with predefined rules, while manual allocation is conducted exclusively by a human Allocator. Semi-manual allocation combines both approaches [5]. The degree of automation affects the flexibility of the applied allocation strategy. Two categories are distinguished: static and dynamic allocation. Static allocation follows predefined rules, limiting the ability to respond to actual traffic situations [7,13]. Dynamic allocation is conducted based on predicted complexity or actual complexity through constant recalculation in the live system [14,15]. Several master’s theses have analyzed initial approaches to allocation strategies that reduce controller workload. One study found that cluster-based, flow-based, and conflict-based allocations had a positive effect on workload reduction in Hungarian Flight Centric ATC airspace. However, only flow-based allocation reduced controllers’ workload in restricted airspace [16]. Another study investigated trajectory-based allocation in a free-route environment for European-wide Flight Centric ATC airspace. This approach was found to be less useful in large airspaces with free-route environments, as aircraft often deviate from traditional tracks, making it difficult to allocate them to controllers [17].

A workload-based allocation formula based on [18] was developed for Hungarian Flight Centric ATC airspace, achieving the equal distribution of workload among all controllers. However, this formula did not account for interaction with radar human–machine interfaces or communication between different controllers [15]. Overall, researchers are exploring various approaches to implementing efficient allocation strategies that reduce controller workload in Flight Centric ATC systems. The development of more sophisticated allocation methods and the consideration of factors like automation and flexibility will be crucial for achieving optimal air traffic management.

3. General Allocation System Set-Up

The first approaches to allocation in the Flight Centric ATC concept show that there are many possibilities for the allocation of aircraft to controllers. However, initial validation studies have also shown that many concept approaches can only be used for certain airspaces and are not universally applicable. Furthermore, no analyses were carried out to determine which tasks the Allocation Center should perform in addition to the actual allocation. In order to be able to implement the Flight Centric ATC concept on an individual basis, a theoretical framework for a possible Allocation Center is described below, which enables the relevant requirements for the seamless implementation of the concept to be achieved. To this end, a set of five high-level needs was defined.

3.1. The Allocation Center Should Be Deployed Regardless of the Airspace Under Consideration, Its Complexity, and the Number of Aircraft and Controllers and Independent of the Controller Team Composition Being Used

It is crucial that the Allocation Center to be developed can be used independently of the external factors of the Flight Centric ATC concept in order to avoid additional development, validation, and implementation costs, which have a negative impact on the cost–benefit analysis of the overall concept. Furthermore, one of the main objectives of the Single European Sky ATM Research Programme is to achieve the highest possible interoperability in European air traffic [19]. In order to develop such a universally usable Allocation Center, it is particularly important to ensure a suitable degree of automation of allocation and the flexibility of the applied allocation strategy. It is important to consider that the different potential levels of automation of an Allocation Center do not automatically align with the corresponding flexibility of the allocation strategies. While all levels of automation can be applied within a static allocation framework, dynamic allocation is only possible with semi-manual and automatic allocation. To fulfill the requirement of the individual use of the Allocation Center, only semi-manual and automatic allocation are suitable, as manual allocation restricts the use of the system excessively.

3.2. The Allocation Center Should Be Highly Automated but Also Enable a Flexible and Individual Reaction to the Current Traffic Situation

The more automated the Allocation Center operates, the more likely the primary goal of achieving the even distribution of aircraft among controllers based on their workload is accomplished. Any human intervention in this process leads to a deviation from this objective or requires explicit system support. However, this also results in the system becoming less adaptable to individual situations as human involvement in the decision-making process decreases. Accordingly, combined semi-manual allocation, where the system’s decisions are monitored and possibly confirmed by a human, is the preferred option for achieving this high-level need. Here, the role of the Allocator is similar to the role of the Supervisor. The tasks of the Supervisor are not standardized, as there is no specific definition in the ICAO SARPs. Therefore, the tasks vary depending on the area control center (ACC), but in general, the Supervisor is responsible for the following tasks, among others [20]:

• Shift rostering;
• Monthly rostering;
• Sector configuration;
• Assigning a person to relieve a controller who has just experienced an emergency situation.

While some tasks, such as rostering, will be adapted in the future due to the new concept but remain in a modified form, other tasks, such as selecting the sector configuration, will be eliminated entirely. This will lead to a reduction in the Supervisor’s tasks so that they may be able to take over the tasks of the Allocator. A new position in the area control center would not be required; instead, the new role of the Allocator will be defined, which the Supervisor will take over in addition to their current role. In order to achieve the required flexible use of the Allocation Center, it is necessary to enable the opening and closing of new controller working positions in order to be able to react effectively to the traffic volume. Although this is similar to the opening and closing of sectors in the current system, it is much more flexible with the new concept. The possibility of re-allocating aircraft that have already been allocated to a Flight Centric ATC controller and are already within the Flight Centric ATC area is also essential for maintaining the workload balance between controllers even in unpredictable situations. Although one of the objectives of this concept is to reduce the number of frequency changes by no longer requiring permanent switches between sectors and thus controllers, preventing this completely would make the concept too rigid and prone to failure.

Regardless of the dynamic allocation design, it must be ensured at any time that each aircraft within the Flight Centric ATC airspace has a distinct allocation. This means that each aircraft is allocated to exactly one controller. It must be ensured that there is no double allocation and that no aircraft remains unallocated.

3.3. The Allocation Strategy Should Ensure the Most Even Possible Workload-Based Distribution of Aircraft to Controllers; Additional Allocation Approaches That Reduce the Overall Workload Should Also Be Considered

In addition to the overall structure, the applied allocation strategy also has a key impact on the Allocation Center. As described in the first point, this is crucial for the universal usage of the Allocation Center. In order to achieve the even distribution of the overall workload among all controllers working in the Flight Centric ATC airspace, a workload-oriented allocation of aircraft as a basic allocation is essential. In this context, the mathematical calculation of the workload can be used as a method to determine this, similar to the calculation in today’s sectorized airspace for determining the sector workload. A distinction has to be made between two formulas for calculating the workload: the pre-allocation formula, which calculates the expected workload for a specific period based on flight plan data, and the dynamic workload formula, which calculates the workload at the present time during active operations based on flight plan data and current trajectory data. The actual allocation of aircraft to controllers is based on this dynamic workload formula, while pre-allocation is only used for the rostering of controllers and pre-planning of the opening and closure of controller working positions (CWPs). An initial workload calculation formula was already developed in [21]. In this case, the total workload that an aircraft can generate results from Conflict Detection and Resolution, aircraft maneuvering, aircraft entering the FCA area from lower airspace, aircraft leaving FCA to a lower sector, first/last contact, coordination with other controllers, monitoring activities, and HMI handling activities. In addition to such a workload-based allocation, which primarily serves to distribute the workload evenly among all controllers, additional allocation strategies can be used to reduce the general workload generated by individual aircraft. One possibility is the allocation of aircraft based on their on-board equipment level. The aim here is to allocate aircraft of the same equipment level to a controller in order to avoid additional workload caused by the permanent change in the type of communication with the aircraft due to different equipment levels. This is described in more detail in the following section.

3.4. The Allocation Center Should Be Applicable Long-Term and Should Consider Corresponding Upcoming Technologies in Order to Easily Implement New Allocation Approaches

In today’s sectorized airspace, communication between air traffic controllers and pilots predominantly relies on voice radios. This has the drawback of generating a relatively high workload on both sides. Additionally, voice radio communication lacks the capability for multiple parties to communicate in parallel, requiring controllers to handle tasks sequentially rather than simultaneously. Communication through a data link addresses this limitation by allowing requests or instructions to be transmitted and received in parallel. There are two types of data links suitable for integration into the Flight Centric ATC concept: Controller–Pilot Data Link Communications (CPDLCs) and L-band Digital Aeronautical Communications System (LDACS). CPDLC currently facilitates the exchange of non-safety-critical information, such as route clearances, weather updates, and administrative queries, in a text-based format between air traffic controllers and pilots. However, its drawbacks include extended transmission durations and a high failure rate. Additionally, the size of transferable data is limited, allowing only individual instructions, i.e., no sequences of instructions or entire trajectories, to be transmitted [22]. Conversely, LDACS represents a novel form of data link not yet in operation. It is designed to enable higher data rates, improved reliability and safety, and a more efficient utilization of available spectra [23,24]. The improvement or introduction of such data link connections enables a better allocation of aircraft to controllers. Initial analyses have revealed that controllers with a dedicated allocation of aircraft equipped with LDACS experience a significantly lower workload than controllers controlling only aircraft equipped with CPDLC or controllers controlling aircraft without a data link [25]. Accordingly, the Allocation Center should consider such types of allocation, even if the corresponding technologies have not yet been implemented in some cases. It should also be taken into account here that LDACS enables complete communication via data link under normal situations so that with a correspondingly higher level of automation of the overall concept, such aircraft would only need to be monitored secondarily by a controller, although it must be ensured at any time that the human controller can intervene in the decisions of the system, including the use of radiotelephony. The re-allocation of these aircraft from the machine to the controller must be ensured. It must also be clearly visualized at any time for all controllers whether an aircraft is primarily under the control of a human ATCO or the system.

3.5. The Allocation Center Is Intended to Support the Rostering of Controllers, Which Differs from Today’s Sector-Based Rostering as It Is More Closely Linked to the Total Traffic Volume

The previous sections primarily dealt with the allocation of aircraft to controllers and the general structure of the Allocation Center. However, the Allocation Center is more than just the pure aircraft–controller allocation. It also includes the general implementation of controller rostering and the associated planning of the opening and closing of controller workstations based on the actual traffic volume. A distinction must be made between monthly rostering and shift rostering. Both types of rostering change as a result of the introduction of the Flight Centric ATC concept, but monthly rostering has no impact on the Allocation Center, as this continues to take place in the pre-planning phase. In contrast, shift rostering, which is one of the tasks of the Supervisor, has a decisive influence on the design of the Allocation Center.

Nowadays, the Supervisor is responsible for shift rostering, which is closely linked to the splitting and merging of sectors. The Supervisor not only has to analyze the traffic demand in the individual sectors, but they also have to take into account the corresponding rating of the available controllers, as not every controller has the appropriate rating for each sector of an ACC [26]. With the Flight Centric ATC concept, on the other hand, the traffic volume in a specific area of airspace is no longer the decisive factor. Instead, the actual traffic volume and the traffic volume forecast for the coming hours for the entire Flight Centric ATC area can be used to determine the corresponding demand on controllers. This provides a basis for determining the point in time at which the currently active controllers reach their maximum workload, and a new controller working position needs to be opened accordingly. The same also applies to the closing of a controller working position, whereby the calculation in this case results in the minimum required workload not being reached. The advantage of this compared to the traditional sectorized concept lies in the fact that all controllers in an ACC have a corresponding rating for the entire FCA airspace and no longer just for parts of it [27,28].

To enable the Supervisor to perform this activity, the corresponding human–machine interface, which is connected to the Allocation Center, must be adapted accordingly. This means that the Supervisor requires information on the following aspects at any given time:

• Workload of all controllers based on the current traffic volume;
• Workload of all controllers based on the predicted traffic volume;
• Available backup controllers.

Based on these data, the Supervisor is able to estimate the demand of controllers for the entire shift and can prepare the corresponding opening/closing of controller workstations.

3.6. The High-Level Needs of the Allocation Center

Based on the five high-level needs mentioned above, the operational requirements that consider the following aspects must be fulfilled:

• Universal use of the concept;
• High degree of automation with the possibility of human intervention;
• Unambiguous allocation;
• Possibility of the opening/closing of controller working positions;
• Possibility of re-allocation;
• Even distribution of workload;
• Consideration of different communication methods (Voice, CPDLC, LDACS);
• Support for rostering.

4. Operational Requirements

Based on the high-level needs of an Allocation Center for the Flight Centric ATC concept, operational requirements can be derived. These serve as a basis for further research on the Allocation Center and as a starting point for the final operational requirements for the subsequent implementation of such a center in the future. Accordingly, the following eleven operational requirements are defined for a future Allocation Center in the Flight Centric ATC concept:

• REQ-OPS-01: The Allocation Center, along with its individual components, shall be usable independently of the Flight Centric ATC airspace under consideration.
• REQ-OPS-02: The Allocation Center shall have a high degree of automation, monitored by a human Allocator, while also allowing for human intervention in the system’s decisions at any given time.
• REQ-OPS-03: The Allocation Center shall allow for the automatic or human-initiated re-allocation of all aircraft already allocated and within the Flight Centric ATC area.
• REQ-OPS-04: The re-allocation shall be clearly visualized for the controllers concerned and displayed on the radar HMI with a lead time.
• REQ-OPS-05: The Allocation Center shall enable the automatic or human-initiated opening or closing of controller working positions at any time.
• REQ-OPS-06: The allocation of aircraft to controllers shall be unambiguous at any time. Each aircraft must be allocated to a specific controller before entering the Flight Centric ATC area. Double allocation must be excluded under all circumstances.
• REQ-OPS-07: The visualization of the allocation shall be clear for all controllers and the Supervisor on their respective radar/Supervisor HMI at any time.
• REQ-OPS-08: The utilized allocation strategy shall enable the even distribution of all aircraft to air traffic controllers based on their individual workload.
• REQ-OPS-09: The Allocation Center shall enable the implementation of new types of communication and also take into account the possibility of higher automation, in which the system has primary control over an aircraft under normal conditions.
• REQ-OPS-10: The Allocation Center shall enable the Supervisor to execute shift rostering and react individually to the current traffic situation.
• REQ-OPS-11: The Allocation Center shall provide the Supervisor with all relevant information on current and expected traffic volumes, abnormal situations, e.g., thunderstorms, and the available ATCOs.

5. Conclusions and Next Steps

It turns out that an Allocation Center enables more than just the pure allocation of aircraft to controllers. In order to successfully implement the Flight Centric ATC concept in air traffic control, it is necessary to comprehensively develop and validate the corresponding center and the human–machine interface of the Supervisor. To be able to use the Allocation Center universally, regardless of the geographical characteristics of the considered Flight Centric ATC airspace, it is important to use only those allocation strategies for workload reduction that cannot be influenced by geographical conditions, in addition to a workload-based allocation. This workload-based allocation should be based on mathematical calculation models that are adapted to the new concept but do not fundamentally differ from the proven concepts of workload calculation in the conventional sector-based environment.

To achieve an optimal allocation even in complex airspaces and regardless of the traffic situation (e.g., thunderstorm, degraded-mode situations), it is essential to apply a semi-manual Allocation Center consisting of an allocation system and a human Allocator. In order to respond to individual situations, the Allocator or the system should be given the ability to perform the re-allocation of aircraft while they are within the Flight Centric ATC area. Furthermore, the opening and closing of controller working positions must be possible in order to manage the respective traffic volume. The allocation carried out and the associated responsibility for each individual aircraft, as well as any type of allocation change, for example, due to re-allocation, must be clearly presented to all controllers and the Supervisor at any time.

Future communication capabilities such as CPDLC and LDACS should be considered in the development of suitable allocation approaches, as they pave the way for a highly automated Air Traffic Control System. With appropriate equipment, the workload of individual aircraft can be significantly reduced, and the involvement of controllers in controlling these aircraft may only be necessary in critical situations, as the system can have general control over them in normal operation.

Rostering is a very complex problem in the sectorized airspace, both in the long term (training of new controllers, rating, etc.) and short term (availability of controllers vs. traffic demand). The Flight Centric ATC concept has the decisive advantage that the elimination of sector boundaries and the resulting standardized training of controllers lead to increased flexibility in controller resource planning. This means that all controllers can be deployed flexibly regardless of the specific “sector rating”. It is also no longer necessary to consider the traffic demand in a particular sector, which in the current system leads to the merging or splitting of sectors. Instead, the number of controllers required is solely dependent on the total demand in the entire Flight Centric ATC airspace.

The statements summarized in this paper on the design of an allocation system for the Flight Centric ATC concept must be analyzed in greater depth in corresponding follow-up work. For this purpose, it is recommended to conduct both fast-time and real-time simulations with such an Allocation Center and evaluate corresponding rostering concepts with active controllers and Supervisors.