Air Route Traffic Control Center Research Articles

Machine learning-enhanced aircraft landing scheduling under uncertainties

Aircraft delays lead to safety concerns and financial losses, which can propagate for several hours during extreme scenarios. Developing an efficient landing scheduling method is one of the effective approaches to reducing flight delays and safety concerns. Existing scheduling practices are mostly done by air traffic controllers (ATC) with heuristic rules. This paper proposes a novel machine learning-enhanced methodology for aircraft landing scheduling. Data-driven machine learning (ML) models are proposed to enhance automation and safety. ML enhancement is adopted for both prediction and optimization. First, the flight arrival delay scenarios are analyzed to identify the delay-related factors, where strong multimodal distributions and arrival flight time duration clusters are observed. A multi-stage conditional ML predictor is proposed for improved prediction performance of separation time conditioned on flight events. Next, we propose incorporating the ML predictions as safety constraints of the time-constrained traveling salesman problem formulation. The scheduling problem is then solved with mixed-integer linear programming (MILP). Additionally, uncertainties between successive flights from historical flight recordings and model predictions are included to ensure reliability. We demonstrate the real-world applicability of our method using the flight track and event data from the Sherlock database of the Atlanta Air Route Traffic Control Center (ARTCC ZTL). The case studies provide evidence that the proposed method is capable of reducing the total landing time by an average of 17.2% across three case studies, when compared to the First-Come-First-Served (FCFS) rule. Unlike the deterministic heuristic FCFS rule, the proposed methodology also considers the uncertainties between aircraft and ensures confidence in the scheduling. Finally, several concluding remarks and future research directions are given. The code used can be retrieved from Link.

Evaluation of Airport Size and Delay Causal Factor Effects on Delay Propagation Dissipation

The International Civil Aviation Organization identifies departure and arrival punctuality as on-time key performance indicators. However, these metrics assume a flight’s delay is a result of either the origin or destination airport, providing limited information on where delay should be mitigated in the U.S. National Airspace System (NAS). This study evaluates the relationship between delay propagation magnitude, delay causal factor, airport size, and charged facility (airport or Air Route Traffic Control Center), to examine if certain delays take longer to dissipate. First, using flights from July 2018, results show that most delay propagation chains originate at large-hub airports. However, these delays were the quickest to recover. Second, this study presents a regression model, predicting total propagated delay using fixed effects based on the weather region where the original delay occurred. Each additional flight affected by downstream delay adds 18.7 min on average to total arrival delay in a propagation chain. Additionally, if weather was the original causal factor, total propagated delay increased by 11.6 min compared with non-weather delays. Lastly, this study compares delay propagation in July 2018 and July 2020. Results show uneven impacts of the coronavirus disease 2019 (COVID-19) across the large-hub airports. Some of the investigated airports did not witness large improvements in average delay per delayed flight, warranting further research in the future. While delay and delay propagation have not been completely eradicated in the NAS during the COVID-19 pandemic, findings suggest that both have significantly declined on average.

A fuel savings and benefit analysis of reducing separation standards in the oceanic airspace managed by the New York Air Route Traffic Control Center

New or improved satellite-based technologies are being introduced to improve the surveillance and communication capabilities in remote airspace. We study the benefits of reducing the separation standards among flights in the oceanic airspace managed by the New York Air Route Traffic Control Center (New York Oceanic) based on these technologies. We develop a model that simulates the activities of aircraft, pilots, and air traffic controllers (ATC) at the microscopic level to study the benefits of doing so in 2020 and 2025. With pessimistic assumptions on the reduced separation standards, the system-wide fuel savings within New York Oceanic are about (in million gallons) 2.25 in 2020 and 3.21 in 2025. After excluding additional variable cost, the monetary value of the fuel savings is about (in million 2018 US dollars) 3.65 and 6.38, respectively. The fuel benefits are more significant for aircraft with light or medium maximum takeoff weight. Some determinants of the workload of ATC and pilots can reduce by about 10% to 20%. With optimistic assumptions on the reduced standards, the corresponding statistics are about 2 to 3 times as high. This study can be used, for example, by air traffic control agencies to conduct benefit-cost analyses of adopting new/improved technologies, by airlines to develop strategies to make the best use of satellite services, and by satellite service providers to design service charging schemes and conduct market analysis.

Human-in-the-Loop Simulation Experiment of Integrated Arrival/Departure Control Services For NextGen Operational Improvement

The National Airspace System suffers from a reduction in airport throughput or even closure when weather or traffic volume constrains arrival or departure gates to and from the Terminal Radar Approach Control (TRACON) facilities. To alleviate this, the Federal Aviation Administration introduced the Integrated Arrival and Departure Control Services (IADCS) concept that would extend terminal separations and procedures to the adjacent Air Route Traffic Control Center (ARTCC) sectors for more flexible traffic. It proposes the use of resectorization, bidirectional gates, bidirectional routes, and Air Traffic Control assigned routes. We evaluated them in the human-in-the-loop high-fidelity experiment and collected objective and subjective data. Our results clearly showed that all IADCS procedures except the vertically separated/bidirectional gate procedure were more effective than the Baseline condition that controllers currently use in the field. We conjecture the vertically separated/bidirectional gate procedure requires more complex perceptual and cognitive processes than the other procedures.

Human Factors Evaluation of Conflict Resolution Advisories for Air Traffic Control in the En Route Domain

In this paper, we report results of a human-in-the-loop simulation experiment that evaluated how Conflict Resolution Advisories (CRA) affected en route air traffic controllers’ performance. Twelve current en route Certified Professional Controllers from Air Route Traffic Control Centers (ARTCCs) participated in the experiment. Results showed that controllers used CRA menus significantly more often than Baseline menus. They also spent more time interacting with the CRA menus than with the Baseline menus. Most of the participants’ subjective ratings favored the CRA, but they also pointed out a few features to be improved.

Sequence Assignment Model for the Flight Conflict Resolution Problem

Each and every day, certain air route traffic control centers receive a large number of requested flight plans, each detailing the level, track, and entry time for the flight to enter the oceanic airspace. Because each airline independently optimizes its own flight plans, it is common that the requested flight plans result in an unbalanced usage of level and track capacities and a conflicting schedule that violates the Federal Aviation Administration (FAA) safety standards. In such a situation, the flight conflict resolution problem is to provide a schedule that minimizes the total penalty cost of delay, level change, and track change while maintaining the FAA separation standards. In this paper, we propose a flight sequence assignment model (FSAM) that selects an optimal set of flight sequences to minimize the total penalty cost. We develop a column generation framework with a bilinear pricing subproblem to solve the linear relaxation of the FSAM, and use a branch-and-price method to find the integer optimal solution. The proposed model is tested on 10 simulated instances randomly constructed based on a real data set. The results show that the FSAM outperforms all other methods in all test instances. We also extend the FSAM to incorporate equity of flight plans among airlines. The computational results show that the equity constraints marginally increase the overall schedule disruption.

Development of a Probabilistic Convective Weather Forecast Threshold Parameter for Flight-Routing Decisions

Abstract This paper presents a method for determining a threshold value of probabilistic convective weather forecast data. By synchronizing air traffic data and an experimental probabilistic convective weather forecast product, it was observed that aircraft avoid areas of specific forecasted probability. Both intensity and echo top of the forecasted weather were synchronized with air traffic data to derive the probability threshold parameter. This value can be used by dispatchers for flight planning and by air traffic managers to reroute streams of aircraft around convective cells. The main contribution of this paper is to provide a method to compute the probability threshold parameters using a specific experimental probabilistic convective forecast product providing hourly guidance up to 6 h. Air traffic and weather data for a 4-month period during the summer of 2007 were used to compute the parameters for the continental United States. The results are shown for different altitudes, times of day, aircraft types, and airspace users. Threshold values for each of the 20 Air Route Traffic Control Centers were also computed. Additional details are presented for seven high-altitude sectors in the Fort Worth, Texas, center. For the analysis reported here, flight intent was not considered and no assessment of flight deviation was conducted since only aircraft tracks were used.

Framework for High-Density-Area Departure and Arrival Traffic Management

An operational concept is proposed to improve high-density-area departure and arrival traffic management that specifically accounts for complications arising in metroplex operations where multiple airports are located in close proximity. In the proposed concept, the roles and responsibilities are redistributed among the traffic management coordinators in different facilities, which include the air route traffic control center, terminal radar approach control, and the airport traffic control tower. The redistribution of roles and responsibilities facilitates improved decision-making capabilities, thereby increasing safe, efficient, and stable operation of departure and arrival traffic in the Next Generation Air Transportation System. This paper proposes a set of functions and capabilities needed to support the roles and responsibilities defined in the proposed concept. The decision support system framework defines three levels of decision making and incorporates an optimization methodology to assist decision makers at the different phases of the decision process. A detailed description of the decomposition and corresponding decision support system structure is presented, and a description of the optimization models is provided. An analysis is performed on a realistic traffic example to demonstrate the optimization model and illustrate the concept.

Collaborative Air Traffic Decision Support Based on Web-based Intelligent Argumentation

Collaborative Decision Making (CDM) is a process of reaching consensus on a potential solution of an issue through the evaluation of the different possible alternatives. The web-based intelligent computational argumen- tation system allows concerned decision making agents to post their arguments on different alternatives, assign degree of strengths to their arguments and identify the most favorable alternative using our system over the internet. Agents are a group of people who participate in the argumentation process for the collaborative decision making process. Our system resolves the conflicts through intelligent argumentation and captures the rationale of the agents from their arguments. The exchange of information among the agents in the form of arguments helps them present their views and opinions and drives the group towards collective intelligence. In this article, we present an approach on how the intelligent argumentation based collaborative decision support system can facilitate the resolution of conflicts in air traffic management. It could enhance the Ground Delay Program (GDP) and help the Air Traffic Control System Com- mand Center (ATCSCC) to take a better decision depending on the argumentation of Air Route Traffic Control Centers (ARTCC) and agents from different airlines.

Dynamic Airspace Configuration Using Approximate Dynamic Programming

On the basis of weather and high traffic, the Next Generation Air Transportation System envisions an airspace that is adaptable, flexible, controller friendly, and dynamic. Sector geometries, developed with average traffic patterns, have remained structurally static with occasional changes in geometry due to limited forming of sectors. Dynamic airspace configuration aims at migrating from a rigid to a more flexible airspace structure. Efficient management of airspace capacity is important to ensure safe and systematic operation of the U.S. National Airspace System and maximum benefit to stakeholders. The primary initiative is to strike a balance between airspace capacity and air traffic demand. Imbalances in capacity and demand are resolved by initiatives such as the ground delay program and rerouting, often resulting in systemwide delays. This paper, a proof of concept for the dynamic programming approach to dynamic airspace configuration by static forming of sectors, addresses static forming of sectors by partitioning airspace according to controller workload. The paper applies the dynamic programming technique to generate sectors in the Fort Worth, Texas, Air Route Traffic Control Center; compares it with current sectors; and lays a foundation for future work. Initial results of the dynamic programming methodology are promising in terms of sector shapes and the number of sectors that are comparable to current operations.

Are There Standard Conflict Resolution Maneuvers in Air Traffic Control?

We analyzed samples of aircraft track data involving conflict alerts and subsequent resolution maneuvers from five U.S. air route traffic control centers. Vertical conflict resolution maneuvers were used in the majority of the cases examined. Within the vertical dimension, reductions of current vertical change (climb or descent) were collectively the most frequent resolution maneuver type, but descents were about three times as frequent as climbs. Three contingency analyses were performed to test for independence between conflict geometries and subsequent resolution maneuvers. The results show that conflict resolution maneuvers do not seem to be independent from conflict geometries. Additional analyses on conflict angles and vertical speed differences between aircraft as continuous variables revealed disproportionately fewer turns on crossing courses than on opposite or same courses. These results provide critical baseline information for understanding just how large departures from the current ingrained practices the advent of the NextGen air traffic management infrastructure may bring about, particularly in extremely safety-critical tasks such as airborne conflict resolutions. Pilots' conflict resolution maneuver preferences have received some attention, but corresponding research on air traffic controllers' practices is almost nonexistent. Moreover, our results emerged from objective data through objective analyses. To our knowledge, this is the first time analyses like these have been applied to data like these.

Algorithms for Combining Airspace Sectors

A heuristic algorithm for combining under-utilized airspace sectors to conserve air traffic control resources is described and analyzed. Simulations and analysis using historical air traffic data and operational sector combination data suggest that systematically combining under-utilized sectors can lead to fewer sectors and therefore a more efficient utilization of resources. Currently, sector combinations are restricted to occur within groups of sectors called areas of specialization. A second heuristic algorithm is proposed that defines new groups of sectors that may be combined. These new groups allow more sector combinations and could be building blocks for new areas of specialization. An analysis of the new groups of sectors suggests that they allow for more frequent sector combinations than existing areas of specialization, and therefore even further efficiencies. Feedback from managers at the Cleveland Air Route Traffic Control Center suggests that both of these algorithms would be useful in Center operations.

Convex Formulations of Air Traffic Flow Optimization Problems

The problem of regulating air traffic in the en route airspace of the National Airspace System is studied using a Eulerian network model to describe air traffic flow. The evolution of traffic on each edge of the network is modeled by a modified Lighthill-Whitham-Richards partial differential equation. The equation is transformed with a variable change, which makes it linear and enables us to use linear finite difference schemes to discretize the problem. We pose the problem of optimal traffic flow regulation as a continuous optimization program in which the partial differential equation appears in the constraints. We propose a discrete formulation of this problem, which makes all constraints (the discretized partial differential equations, boundary, and initial conditions) linear. Corresponding linear programming and quadratic programming based solutions to this convex optimization program yield globally optimal solutions to various air traffic management objectives. The proposed method is applied to the maximization of aircraft arrivals and minimization of delays in the arrival airspace due to exogenous capacity reductions. The corresponding linear and quadratic programs are solved numerically using CPLEX for a benchmark scenario in the Oakland Air Route Traffic Control Center. Several computational aspects of the method are assessed-in particular, accuracy of the numerical discretization, computational time, and storage space required by the method.

The Relationship between Aircraft Count and Controller Workload in Different En Route Workstation Systems

Some air traffic control researchers have reported that the relationship between aircraft count in the sector and controller workload is nonlinear. Other researchers have reported it linear. However, they did not manipulate the number of aircraft separately in their experiments. In our experiment, we gradually increased traffic levels within each experimental run. Our controllers used three different workstation systems with and without data communication. The linear and power models described the relationship well. The proportion of workload variance explained by aircraft count was well above 50%. With these models, we predict controllers could handle about four (13%) more aircraft with data communication than without it with the Display System Replacement (DSR) system that is currently used at Air Route Traffic Control Centers (ARTCCs).

The Design of a Distributed Scheduling System for Multi-Center Time-Based Metering of Air Traffic into Congested Resources

A distributed scheduling system has been developed and implemented within NASA’s Multi-center Traffic Management Advisor (McTMA) proto-type research system. This paper describes the design and operation of the distributed scheduling system, its testing, and potential future use. The distributed scheduler is a loosely-coupled network of schedulers, each operating over short distances to deconflict aircraft merging into local points of congestion, while informing all other networked schedulers of its constraints. This architecture coordinates aircraft arrivals into congested resources by allocating delay over long distances while being robust to uncertainties in long-range estimates of arrival times. The system reduced airborne delay by 20–42% (without a comparable increase in ground delay), reduced airborne holding by 90%, reduced vectoring, and resulted in a shorter final approach in field tests at four FAA Air Route Traffic Control Centers, the Philadelphia Terminal Radar Approach Control facility, and the FAA’s Air Traffic Control System Command Center. The FAA is currently incorporating technology from the distributed scheduler into its nationwide deployment of the Traffic Management Advisor system.

Concept and Laboratory Analysis of Trajectory-Based Automation for Separation Assurance

An operating concept and a laboratory analysis methodology were developed and tested to examine how four-dimensional trajectory analysis methods could support higher levels of automation for separation assurance in the National Airspace System. Real-time simulations were conducted in which a human controller generated conflict resolution trajectories using an automated trial plan trajectory generation and analysis function, but only in response to conflicts detected and displayed by an automatic conflict detection function. Objective metrics were developed to compare aircraft separation characteristics and flying time efficiency under automated operations with that of today’s operations using common airspace and common traffic scenarios. Simulations were based on recorded air traffic data from the Fort Worth Air Route Traffic Control Center and were conducted using today’s and nearly two-times today’s traffic levels. The results suggest that a single controller using trajectory-based automation and data link communication of control clearances to aircraft could manage substantially more traffic than under today’s conditions, and with improved route efficiency while maintaining separation. The simulation and analysis capability provides a basis for further analysis of semi-automated, or fully automated, separation assurance concepts.

Comparison of the performance of four Eulerian network flow models for strategic air traffic management

Four Eulerian network models are implemented to model high altitude air traffic flow. Three of the models use the framework of discrete time dynamical systems, while the fourth consists of a network of partial differential equations. The construction of these models is done using one year of air traffic data. The four models are applied to high altitude traffic for six Air Route Traffic Control Centers in the National Airspace System and surrounding airspace. Simulations are carried out for a full day of data for each of the models, to assess their predictive capabilities. The models’ predictions are compared to the recorded flight data. Several error metrics are used to characterize the relative accuracy of the models. The efficiency of the respective models is also compared in terms of computational time and memory requirements for the scenarios of interest. Control strategies are designed and implemented on similar benchmark scenarios for two of the models. They use techniques such as adjoint-based optimization, as well as mixed integer linear programming. A discussion of the four models’ structural differences explains why one model may outperform another.

Predeparture Uncertainty and Prediction Performance in Collaborative Routing Coordination Tools

The collaborative routing coordination tools concept demonstration and evaluation prototype is a tool to help the Federal Aviation Administration detect traffic flow problems in advance, generate reroute-based problem resolutions, and evaluate the resolution strategies. The tool does this by modeling four-dimensional aircraft trajectories and using them to predict demand for sector usage. The predictions are affected by various sources of uncertainty in the national airspace system. Therefore, it is useful to understand the contribution of this uncertainty to the tool's predictive error. A methodology was developed and used to assess the impact of predeparture uncertainty on the prototype tool's prediction performance. The tool's trajectory modeling and sector load prediction are described, the methodology used to assess the tool's prediction performance is presented, and the results of the predeparture uncertainty analysis are discussed. The analysis is based on tool predictions made for two air route traffic control centers. Tool runs were made against recordings of actual air traffic data on two good weather and two bad weather days. Analysis of run data showed that predeparture uncertainty is a greater contributor to sector prediction error than en route uncertainty. Because predeparture uncertainty is reflected in departure time estimate, research is recommended to identify methods to improve the accuracy of this estimate.

Economic Benefits of Increased En Route Sector Capacity from Controller-Pilot Data Link Communications

The En Route Controller-Pilot Data Link Communications (CPDLC) program is designed to enable digital data communications between controllers and pilots. CPDLC Build 1A has been proposed as the first step in the nationwide rollout of this program to en route airspace by implementation in eight air route traffic control centers. Because CPDLC will automate several routine air-ground communications tasks, one potential benefit is reduced controller workload, which could lead to increased sector capacity and thus reduced flight times for flights traversing busy sectors. To quantify the workload reduction, a comparison has been made of the time currently required to perform communications tasks, as determined from recordings of actual air-ground communications, and the time required to perform the same tasks with CPDLC, as determined by a controller-in-the-loop simulation. This comparison indicated the potential for a significant reduction in the time spent on controller-pilot communications. A model was constructed to relate the reduced controller communications workload to increased sector capacity. The National Airspace System Performance Analysis Capability (NASPAC) model was used to estimate the reduction in flight times associated with the modeled sector capacity increase. NASPAC was run for both an idealized good weather day and a sample actual bad weather day. By 2017, the last year of the projected economic lifetime of CPDLC, average flight time savings from CPDLC-related capacity increases were estimated to exceed 9, 400 min/day.

Ground Delay Programs: Optimizing over the Included Flight Set Based on Distance

The Ground Delay Program (GDP) is an air traffic flow management mechanism used to decrease the rate of in-coming flights into an airport when it is projected that arrival demand will exceed capacity. Under a GDP, a set of flights destined for a single airport is assigned ground delays. In this paper we investigate how the set of flights to which delays are applied is defined. Specifically, we define a “distance-based” GDP as one that only applies to flights whose origin airports are less than a prescribed distance, d, from the destination airport. This approach is different from the current approach which groups origin airports by air route traffic control center jurisdiction and restricts flight based on a center-based tier system.We also investigate methods for setting the parameter d. We describe the two measures currently used to evaluate GDP options, i.e. unrecoverable delay and average delay, and show how to optimize d based on an objective function defined in terms of these parameters. A new GDP measure, unnecessary delay, which represents the expected cost of ground delay that was unnecessarily assigned is defined. It is shown that this measure provides an improvement over unrecoverable delay.