Maxime Gariel

Trajectory Clustering and an Application to Airspace Monitoring

This paper presents a framework aimed at monitoring the behavior of aircraft in a given airspace. Trajectories that constitute typical operations are determined and learned using data-driven methods. Standard procedures are used by air traffic controllers (ATCs) to guide aircraft, ensure the safety of the airspace, and maximize runway occupancy. Even though standard procedures are used by ATCs, control of the aircraft remains with the pilots, leading to large variability in the flight patterns observed. Two methods for identifying typical operations and their variability from recorded radar tracks are presented. This knowledge base is then used to monitor the conformance of current operations against operations previously identified as typical. A tool called AirTrajectoryMiner is presented, aiming at monitoring the instantaneous health of the airspace, in real time. The airspace is “healthy” when all aircraft are flying according to typical operations. A measure of complexity is introduced, measuring the conformance of current flight to typical flight patterns. When an aircraft does not conform, the complexity increases as more attention from ATC is required to ensure safe separation between aircraft.

Data-Based Modeling and Optimization of En Route Traffic

Air traffic management aims at ensuring safe and efficient movement of aircraft in the airspace. With the predicted growth of air transportation, providing traffic flow managers with the tools to support decision-making is essential. These tools should aid in accommodating the air traffic throughput increase while limiting controller workload and ensuring high safety levels. The objective of this paper is to present a methodology to model and simulate traffic in a given portion of the airspace from data under nominal and perturbed conditions. A new framework for en route traffic flow management and airspace health monitoring is developed. It is based on a data-driven approach for air traffic flow modeling using historical data. This large-scale three-dimensional flow network provides valuable insight on airspace complexity. A linear programming formulation for optimizing en route air traffic is proposed. It takes into account a controller task load model based on flow geometry, in order to estimate airspa...

Determining bounds on controller workload rates at an intersection

This paper considers the problem of determining the maximum controller workload associated with two intersecting flows of aircraft according to aircraft arrival rates. Based on the structure of the aircraft arrivals, an optimal control strategy is presented for minimizing the maximum rate of resolutions required to deconflict traffic at the intersection. Bounds on the rate of resolution take into account magnitude constraints in the conflict resolution commands issued to aircraft. The goal of this research is to establish worst-case measures of workload such that air traffic flow managers can route traffic along network structures without exceeding human-controller performance capabilities.

Airspace Statistical Proximity Maps Based on Data-Driven Flow Modeling

This paper presents a new methodology that aims to rapidly generate 3-D sector proximity maps, which indicate the probability of presence of at least one or two aircraft at any given point in a considered sector. The maps are generated using an aircraft flow model driven from historical data. Time-varying flow characteristics such as routes, speed, probability density function of the inter-arrival time between two consecutive aircraft, were determined using ETMS data. The maps are intended to be a predictive tool for traffic flow management in order to anticipate for a given time period how different flows may interact together and to predict which “critical” regions may be subject to possible conflict between aircraft.

Airspace Complexity Estimations Based on Data-Driven Flow Modeling

This paper presents a new methodology that aims to rapidly generate airspace complexity estimations, which are principally based on the probability of presence of at least one or two aircraft at any given point in a considered sector. Three-dimensional complexity maps are generated using an aircraft flow model driven from historical data. Time-varying flow characteristics such as routes, speed, probability density function of the inter-arrival time between two consecutive aircraft, are determined using Enhanced Traffic Management System (ETMS) data. In addition to the flow characteristics, the complexity maps take into account the sector geometrical configuration and the probability of severe weather. From the complexity maps, a scalar estimation of the airspace complexity is proposed. The complexity estimations presented in this paper are intended to be a predictive tool to support traffic flow management, in order to anticipate for a given time period how different flows may interact together. Especially, the 3-D complexity maps allow to predict which “critical” regions may be subject to possible conflict between aircraft or to the presence of aircraft in severe weather area.

A Dynamic I/O Model for TRACON Traffic Management

This work investigates the TRACON flow management around a major airport. Aircraft flows are analyzed through a study of TRACON trajectories records. Rerouting and queuing processes are highlighted and airport characteristics are shown as function of the number of planes in the TRACON. Then, a simple input-output TRACON queuing and landing model is proposed. This model is calibrated and validated using available TRACON data. It reproduces the same phenomenon as the real system. This model is used to show the impact of limiting the number of aircrafts in the TRACON. A limited number of aircraft does not increase delays but reduces the controllers workload and increases safety.

Air traffic optimization on data-driven network flow model

This paper presents a new framework for Traffic Flow Management and Airspace Health Monitoring based on data-driven approach for air traffic flows modeling using historical data. The large-scale 3-dimensional flow network of the Cleveland center airspace provides valuable insight on airspace complexity. A linear formulation of the Traffic Flow Management Problem is proposed, taking into account estimations of controller workload based on flow geometry. Preliminary results for the problem are discussed, pointing out clues for further research.

3D conflict avoidance under uncertainties

This paper presents a 3D conflict avoidance algorithm in the presence of uncertainty. The objective of the algorithm is to ensure the safety of the airspace in the event of a failure in the communication, navigation or surveillance systems. The algorithm minimizes the number of maneuvers required to maintain the safety of the airspace under degraded conditions. Uncertainties are modeled as an increase in the required separation distance between aircraft. A single maneuver for each aircraft is chosen to maintain safe separation. Maneuvers include heading change, speed change and flight level change. Maneuvers are simple to execute and guarantee a conflict-free configuration after execution. Their feasibility is constrained by weather avoidance, sector boundaries and aircraft performance. A Mixed Integer Program is used to determine the set of maneuvers to be executed.

Air Traffic Management complexity maps induced by degradation of Communication, Navigation and Surveillance.

The introduction of new technologies and concepts of operation in the air transportation system is not possible, unless they can be proven not to adversely aect the system operation under not only nominal, but also degraded conditions. In extreme scenarios, degraded operations due to partial or complete technological failures should not endanger system safety. Many past system evolutions, whether ground-based or airborne, have been based on trial-and-error, and system safety was addressed only after a specic event yielded dramatic or near-dramatic consequences. Future system evolutions, however, must leverage available computation, prior knowledge and abstract reasoning to anticipate all possible system degradations and prove that such degradations are graceful and safe. Building upon prior research eorts, this paper is concerned with evaluating the ability for a given airspace structure to handle trac

A Conflict Resolution Algorithm For Reduced Controller Taskload

This paper considers the con ict resolution problem arising from a large number of aircraft traversing a shared airspace. Aircraft are issued lateral and longitudinal shifts commands over their trajectory to resolve potential con icts. Lateral and longitudinal shifts given to each aircraft can be abstracted to represent the heading and speed changes over the trajectories. Through this formulation, the resulting con ict resolution algorithm can be formulated as a mixed-integer linear program solvable in real-time. In this model, the procedure aims to assign to each aircraft a sequence of maneuvers instead of a single one. This new method is based on a centralized algorithm with the goal of minimizing the total number of maneuvers issued by the controller while regularizing the trajectories.