Yoon C. Jung

A Mixed Integer Linear Program for Airport Departure Scheduling

A mixed integer linear program is presented for deterministically scheduling departure aircraft at runways. The method addresses different schemes of managing the departure queuing area by treating it as first -in-first-out queues or as a simple parking area, wher e any available aircraft can take-off irrespective of its relative sequence with others. The method explicitly considers separation criteria between successive departures and also incorporates an optional prioritization scheme using time windows. Multiple objectives pertaining to throughput, system delay and maximum individual delay are used. Results indicate minimizing system delay alone improves throughput over a basic first -come-first-serve rule. Modifications for computational efficiency are also presented in the form of re-formulating certain constraints and defining additional inequalities for better bounds.

Managing departure aircraft release for efficient airport surface operations

This paper presents a model for managing departure aircraft at the spot or gate on the airport surface. The model is applied over two time frames: long term (one hour in future) for collaborative decision making, and short term (immediate) for decisions regarding the release of aircraft. The purpose of the model is to provide the controller a schedule of spot or gate release times optimized for runway utilization. This model was tested in nominal and heavy surface traffic scenarios in a simulated environment, and results indicate average throughput improvement of 10% in high traffic scenarios even with up to two minutes of uncertainty in spot arrival times.

A Concept and Implementation of Optimized Operations of Airport Surface Traffic

This paper presents a new concept of optimized surface operations at busy airports to improve the efficiency of taxi operations, as well as reduce environmental impacts. The suggested system architecture consists of the integration of two decoupled optimization algorithms. The Spot Release Planner provides sequence and timing advisories to tower controllers for releasing departure aircraft into the movement area to reduce taxi delay while achieving maximum throughput. The Runway Scheduler provides take-off sequence and arrival runway crossing sequence to the controllers to maximize the runway usage. The description of a prototype implementation of this integrated decision support tool for the airport control tower controllers is also provided. The prototype decision support tool was evaluated through a human-in-the-loop experiment, where both the Spot Release Planner and Runway Scheduler provided advisories to the Ground and Local Controllers. Initial results indicate the average number of stops made by each departure aircraft in the departure runway queue was reduced by more than half when the controllers were using the advisories, which resulted in reduced taxi times in the departure queue.

Incorporating Active Runway Crossings in Airport Departure Scheduling

A mixed integer linear program is presented for deterministically scheduling departure and arrival aircraft at airport runways. This method addresses different schemes of managing the departure queuing area by treating it as first-in-first-out queues or as a simple par king area where any available aircraft can take-off ir respective of its relative sequence with others. In addition, this method explicitly considers separation criteria between successive aircraft and also incorporates an optional prioritization scheme using time windows. Multiple objectives pertaining to throughput and system delay are used independently. Results indicate improvement over a basic first-come-first-serve rule in both system delay and throughput. Minimizing system delay results in small deviations from optimal throughput, whereas minimizing throughput results in large deviations in system delay. Enhancements for computational efficiency are also presented in the form of reformulating certain constraints and defining additional inequalities for better bounds.

An Integrated Collaborative Decision Making and Tactical Advisory Concept for Airport Surface Operations Management

Surface operations at airports in the US are based on tactical operations, where departure aircraft primarily queue up and wait at the departure runways. There have been attempts to address the resulting inefficiencies with both strategic and tactical tools for metering departure aircraft. This paper presents Spot And Runway Departure Advisor with Collaborative Decision Making (SARDA-CDM): an integrated strategic and tactical system for improving surface operations by metering departure aircraft. SARDA-CDM is the augmentation of ground and local controller advisories through sharing of flight movement and related operations information between airport operators, flight operators and air traffic control at the airport. The goal is to enhance the efficiency of airport surface operations by exchanging information between air traffic control and airline operators, while minimizing adverse effects on stakeholders and passengers. The paper presents the concept of operations for SARDA-CDM, describing both the strategic and tactical components. Then the preliminary results from testing the concept in a real-time automated simulation environment are described. Results indicate benefits such as reduction in taxiing delay and fuel consumption. Further, the preliminary implementation of SARDA-CDM seems robust for two minutes delay in gate push-back times.

A Simulator for Modeling Aircraft Surface Operations at Airports

Rapid prototyping of integrated taxiway and runway schedulers at airports has become highly desirable for Safe and Efficient Surface Operations (SESO) to support the Next Generation Air Transportation System (NextGen). Finding optimal solutions for taxiway and runway scheduling are generally computationally difficult and therefore it is essential to develop fast algorithms that can produce high quality solutions for the scheduling problems. Equally important is to develop a fast-time simulation that can be integrated with these algorithms and can handle high traffic demand in the presence of uncertainties. This paper describes the development of Surface Operations Simulator and Scheduler (SOS) that can facilitate the integration of several scheduling algorithms with the surface simulation of the aircraft. The SOS architecture is used to implement First-Come-First-Served (FCFS) heuristics and a comparison to the average taxi times of actual data from Dallas Fort Worth International (DFW) airport with SOS output is made. The integration of a dynamic programming departure scheduler with FCFS taxiway heuristics demonstrates the modular architecture of SOS and a comparison to the output of pure FCFS heuristics demonstrates the potential system-wide savings in taxi delay over FCFS scheduling.

Effect of Uncertainty on Deterministic Runway Scheduling

Active runway scheduling involves scheduling departures for takeoffs and arrivals for runway crossing subject to numerous constraints. This paper evaluates the effect of uncertainty on a deterministic runway scheduler. The evaluation is done against a firstcome-first-serve scheme. In particular, the sequence from a deterministic scheduler is frozen and the times adjusted to satisfy all separation criteria; this approach is tested against FCFS. The comparison is done for both system performance (throughput and system delay) and predictability, and varying levels of congestion are considered. The modeling of uncertainty is done in two ways: as equal uncertainty in availability at the runway as for all aircraft, and as increasing uncertainty for later aircraft. Results indicate that the deterministic approach consistently performs better than first-come-first-serve in both system performance and predictability.

Spot Release Planner: Efficient Solution for Detailed Airport Surface Traffic Optimization

The Spot Release Planner (SRP) is an algorithm previously developed by the authors to reduce delay and congestion on the airport surface. The algorithm was developed to provide real time advisories to tower controllers. A Human-in-the-loop (HITL) simulation in April 2010 showed that the SRP reduced the average movement area delay of departure aircraft by 64%. The SRP is a two-stage algorithm that considers runway scheduling in the rst stage, and the rest of the ground movement, such as gate pushback and spot release, in the second stage. This decomposition of airport surface scheduling into two stages provides fast computational times and makes the SRP applicable for real-time decision making. However, the two stages also result in the given scheme being a heuristic for solving the complicated airport surface scheduling problem; no guarantees on quality of the obtained solution have been provided. This paper explores the quality of solutions obtained by the SRP and compares them with the optimal solution for airport surface trac. Simulations conducted for the East side of Dallas/Fort Worth International Airport (DFW) show that the SRP solutions are within 14s of the optimal solution for a detailed airport surface planner.

Taxi-Out Time Prediction for Departures at Charlotte Airport Using Machine Learning Techniques

Predicting the taxi-out times of departures accurately is important for improving airport efficiency and takeoff time predictability. In this paper, we attempt to apply machine learning techniques to actual traffic data at Charlotte Douglas International Airport for taxi-out time prediction. To find the key factors affecting aircraft taxi times, surface surveillance data is first analyzed. From this data analysis, several variables, including terminal concourse, spot, runway, departure fix and weight class, are selected for taxi time prediction. Then, various machine learning methods such as linear regression, support vector machines, k-nearest neighbors, random forest, and neural networks model are applied to actual flight data. Different traffic flow and weather conditions at Charlotte airport are also taken into account for more accurate prediction. The taxi-out time prediction results show that linear regression and random forest techniques can provide the most accurate prediction in terms of root-mean-square errors. We also discuss the operational complexity and uncertainties that make it difficult to predict the taxi times accurately.

A Concept of Operations for Trajectory-based Taxi Operations

A harmonized concept of operations (ConOps) for future surface operations that considers the surface management practices and policies that currently exist in the U.S. and Europe was developed by NASA and DLR. The high-level concept vision is to develop the framework for surface traffic scheduling systems that generate conflict-free four-dimensional trajectories (4DTs) for all aircraft on the airport surface and for guidance means on board and on ground to enable the flight crew to adhere to the trajectories. This vision supports the reduction of temporal uncertainty and delays by improving the whole planning chain from gate to runway and vice versa. This paper identifies and describes the necessary functions of this concept and explains the relationships among them. As a result of this activity, challenges arose which are the basis to derive research requirements that are jointly approached by NASA and DLR in the future. The concept is supported through a solid base of results from research already conducted in that area to support the implementation of this concept in the future.