Accuracy of reinforcement learning algorithms for predicting aircraft taxi-out times: A case-study of Tampa Bay departures

Taxi-out delay is a significant portion of the block time of a flight. Uncertainty in taxi-out times reduces predictability of arrival times at the destination. This in turn results in inefficient use of airline resources such as aircraft, crew, and ground personnel. Taxi-out time prediction is also a first step in enabling schedule modifications that would help mitigate congestion and reduce emissions. The dynamically changing operation at the airport makes it difficult to accurately predict taxi-out time. In this paper we investigate the accuracy of taxi out time prediction using a nonparametric reinforcement learning (RL) based method, set in the probabilistic framework of stochastic dynamic programming. A case-study of Tampa International Airport (TPA) shows that on an average, with 93.7% probability, on any given day, our predicted mean taxi-out time for any given quarter, matches the actual mean taxi-out time for the same quarter with a standard error of 1.5 min. Also, for individual flights, the taxi-out time of 81% of them were predicted accurately within a standard error of 2 min. The predictions were done 15 min before gate departure. Gate OUT, wheels OFF, wheels ON, and gate IN (OOOI) data available in the Aviation System Performance Metric (ASPM) database maintained by the Federal Aviation Administration (FAA) was used to model and analyze the problem. The prediction accuracy is high even without the use of detailed track data.     

Estimation of aircraft taxi fuel burn using flight data recorder archives

This paper builds a model for estimating the fuel consumption of a taxiing aircraft using flight data recorder information from operational aircraft. The taxi fuel burn is modeled as a linear function of several potential explanatory variables including the taxi time, number of stops, number of turns and number of acceleration events, and the coefficients are estimated using least-squares regression. The statistical significance of each potential factor is investigated. Our analysis shows that in addition to the taxi time, the number of acceleration events is a significant factor in determining taxi fuel consumption. Since the model parameters are estimated using data from operational aircraft, they provide more accurate estimates of fuel burn than methods that use idealized physical models of fuel consumption based on aircraft velocity profiles, or the baseline fuel consumption estimates provided by the International Civil Aviation Organization.     

Assessing the impact of tactical airport surface operations on airline schedule block time setting

With the growth of air traffic, airport surfaces are congested and air traffic operations are disrupted by the formation of bottlenecks on the surface. Hence, improving the efficiency and predictability of airport surface operations is not only a key goal of NASA’s initiatives in Integrated Arrival/Departure/Surface (IADS) operations, but also has been recognized as a critical aspect of the FAA NextGEN implementation plan. While a number of tactical initiatives have been shown to be effective in improving airport surface operations from a service provider’s perspective, their impacts on airlines’ scheduled block time (SBT) setting, which has been found to have direct impact on airlines’ on-time performance and operating cost, have received little attention. In this paper, we assess this impact using an econometric model of airline SBT combined with a before/after analysis of the implementation of surface congestion management (SCM) at John F. Kennedy International Airport (JFK) in 2010. Since airlines do not consider gate delay in setting SBT, we find that reduction in taxi-out time variability resulting from SCM leads to more predictable taxi-out times and thus decreases in SBT. The JFK SCM implementation is used as a case study to validate model prediction performance. The observed SBT decrease between 2009 and 2011 at JFK is 4.8 min and our model predicts a 4.2 min decrease. In addition, Charlotte Douglas International Airport (CLT) is used as an example to demonstrate how different surface operations improvements scenarios can be evaluated in terms of SBT reduction.     

Heuristic search for the coupled runway sequencing and taxiway routing problem

This paper presents the first local search heuristic for the coupled runway sequencing (arrival & departure) and taxiway routing problems, based on the receding horizon (RH) scheme that takes into account the dynamic nature of the problem. As test case, we use Manchester Airport, the third busiest airport in the UK. From the ground movement perspective, the airport layout requires that departing aircraft taxi across the arrivals runway. This makes it impossible to separate arrival from departure sequencing in practice. Operationally, interactions between aircraft on the taxiways could prevent aircraft from taking off from, or landing on, runways during the slots assigned to them by an algorithm optimizing runway use alone. We thus consider the interactions between arrival and departure aircraft on the airport surface. Compared to sequentially optimized solutions, the results obtained with our approach indicate a significant decrease in the taxiway routing delay, with generally no loss in performance in terms of the sequencing delay for a regular day of operations. Another benefit of such a simultaneous optimization approach is the possibility of holding aircraft at the stands for longer, without the engines running. This significantly reduces the fuel burn, as well as bottlenecks and traffic congestion during peak hours that are often the cause of flight delays due to the limited amount of airport surface space available. Given that the maximum computing time per horizon is around 95 s, real-time operation might be practical with increased computing power.     

Maximizing airborne delay at no extra fuel cost by means of linear holding

This paper introduces a linear holding strategy based on prior works on cruise speed reduction, aimed at performing airborne delay at no extra fuel cost, as a complementary strategy to current ground and airborne holding strategies. Firstly, the equivalent speed concept is extended to climb and descent phases through an analysis of fuel consumption and speed from aircraft performance data. This gives an insight of the feasibility to implement the concept, differentiating the case where the cruise flight level initially requested is kept and the case where it can be changed before departure in order to maximize the linear holding time. Illustrative examples are given, where typical flights are simulated using an optimal trajectory generation tool where linear holding is maximized while keeping constant the initially planned fuel. Finally, the effects of linear holding are thoroughly assessed in terms of the vertical trajectory profiles, range of feasible speed intervals and trade-offs between fuel and time. Results show that the airborne delay increases significantly with nearly 3-fold time for short-haul flights and 2-fold for mid-hauls to the cases in prior works.     

Predictability impacts of airport surface automation

Past evaluations of airport surface operations automation technologies have focused on capacity utilization, delay mitigation and fuel efficiency impacts. Predictability, while recognized as an important operational performance goal, has received little attention. One reason could be that applicable predictability metrics have not been developed in the context of airport surface operations management. This research fills the gap by proposing metrics for predictability performance evaluation. Using results from a SARDA human-in-the-loop simulation conducted at NASA Ames’ Future Flight Central, we present a comprehensive assessment of the predictability impacts of airport surface automation. A wide range of the impacts is considered, which includes variability in taxi-out time, predictability of take-off time and take-off sequence, entropy of the airfield state, and perceived predictability from users.     

Time to burn: Flight delay,terminal efficiency,and fuel consumption in the National Airspace System

Improved Air Traffic Management (ATM) leading to reduced en route and gate delay, greater predictability in flight planning, and reduced terminal inefficiencies has a role to play in reducing aviation fuel consumption. Air navigation service providers are working to quantify this role to help prioritize and justify ATM modernization efforts. In the following study we analyze actual flight-level fuel consumption data reported by a major U.S. based airline to study the possible fuel savings from ATM improvements that allow flights to better adhere to their planned trajectories both en route and in the terminal area. To do so we isolate the contribution of airborne delay, departure delay, excess planned flight time, and terminal area inefficiencies on fuel consumption using econometric techniques. The model results indicate that, for two commonly operated aircraft types, the system-wide averages of flight fuel consumption attributed to ATM delay and terminal inefficiencies are 1.0–1.5% and 1.5–4.5%, respectively. We quantify the fuel impact of predicted delay to be 10–20% that of unanticipated delay, reinforcing the role of flight plan predictability in reducing fuel consumption. We rank terminal areas by quantifying a Terminal Inefficiency metric based on the variation in terminal area fuel consumed across flights. Our results help prioritize ATM modernization investments by quantifying the trade-offs in planned and unplanned delays and identifying terminal areas with high potential for improvement.     

Demonstration of reduced airport congestion through pushback rate control

Airport surface congestion results in significant increases in taxi times, fuel burn and emissions at major airports. This paper describes the field tests of a congestion control strategy at Boston Logan International Airport. The approach determines a suggested rate to meter pushbacks from the gate, in order to prevent the airport surface from entering congested states and to reduce the time that flights spend with engines on while taxiing to the runway. The field trials demonstrated that significant benefits were achievable through such a strategy: during eight four-hour tests conducted during August and September 2010, fuel use was reduced by an estimated 12,250–14,500 kg (4000–4700 US gallons), while aircraft gate pushback times were increased by an average of only 4.4 min for the 247 flights that were held at the gate.     

Conflict‐free time‐based trajectory planning for aircraft taxi automation with refined taxiway modeling

To mitigate airport congestion caused by increasing air traffic demand, the trajectory‐based surface operations concept has been proposed to improve surface movement efficiency while maintaining safety. It utilizes decision support tools to provide optimized time‐based trajectories for each aircraft and uses automation systems to guide surface movements and monitor their conformance with assigned trajectories. Whether the time‐based trajectories can be effectively followed so that the expected benefits can be guaranteed depends firstly on whether these trajectories are realistic. So, this paper first deals with the modeling biases of the network model typically used for taxi trajectory planning via refined taxiway modeling. Then it presents a zone control‐based dynamic routing and timing algorithm upon the refined taxiway model to find the shortest time taxi route and timings for an aircraft. Finally, the presented algorithm is integrated with a sequential planning framework to continuously decide taxi routes and timings. Experimental results demonstrate that the solution time for an aircraft can be steadily around a few milliseconds with timely cleaning of expired time windows, showing potential for real‐time decision support applications. The results also show the advantages of the proposed methodology over existing approaches. Copyright © 2015 John Wiley & Sons, Ltd.     

Tarmac delay policies: A passenger-centric analysis

In this paper, we analyze the effectiveness of the 2010 Tarmac Delay Rule from a passenger-centric point of view. The Tarmac Delay Rule stipulates that aircraft lift-off, or an opportunity for passengers to deplane, must occur no later than 3 h after the cabin door closure at the gate of the departure airport; and that an opportunity for passengers to deplane must occur no later than 3 h after the touchdown at the arrival airport. The Tarmac Delay Rule aims to protect enplaned passengers on commercial aircraft from excessively long delays on the tarmac upon taxi-out or taxi-in, and monetarily penalizes airlines that violate the stipulated 3-h tarmac time limit. Comparing the actual flight schedule and delay data after the Tarmac Delay Rule was in effect with that before, we find that the Rule has been highly effective in reducing the frequency of occurrence of long tarmac times. However, another significant effect of the rule has been the rise in flight cancellation rates. Cancellations result in passengers requiring rebooking, and often lead to extensive delay in reaching their final destinations. Using an algorithm to estimate passenger delay, we quantify delays to passengers in 2007, before the Tarmac Delay Rule was enacted, and compare these delays to those estimated for hypothetical scenarios with the Tarmac Delay Rule in effect for that same year. Our delay estimates are calculated using U.S. Department of Transportation data from 2007. Through our results and several sensitivity analyses, we show that the overall impact of the current Tarmac Delay Rule is a significant increase in passenger delays, especially for passengers scheduled to travel on the flights which are at risk of long tarmac delays. We evaluate the impacts on passengers of a number of rule variations, including changes to the maximum time on the tarmac, and variations in that maximum by time-of-day. Through extensive scenario analyses, we conclude that a better balance between the conflicting objectives of reducing the frequency of long tarmac times and reducing total passenger delays can be achieved through a modified version of the existing rule. This modified version involves increasing the tarmac time limit to 3.5 h and only applying the rule to flights with planned departure times before 5pm. Finally, in order to implement the Rule more effectively, we suggest the tarmac time limit to be defined in terms of the time when the aircraft begin returning to the gate instead of being defined in terms of the time when passengers are allowed to deplane.     

Operating cost based cruise speed reduction for ground delay programs: Effect of scope length

Ground delay programs typically involve the delaying of aircraft that are departing from origin airports within some set distance of a capacity constrained destination airport. Long haul flights are not delayed in this way. A trade-off exists when fixing the distance parameter: increasing the ‘scope’ distributes delay among more aircraft and may reduce airborne holding delay but could also result in unnecessary delay in the (frequently observed) case of early program cancellation. In order to overcome part of this drawback, a fuel based cruise speed reduction strategy aimed at realizing airborne delay, was suggested by the authors in previous publications. By flying slower, at a specific speed, aircraft that are airborne can recover part of their initially assigned delay without incurring extra fuel consumption if the ground delay program is canceled before planned. In this paper, the effect of the scope of the program is assessed when applying this strategy. A case study is presented by analyzing all the ground delay programs that took place at San Francisco, Newark Liberty and Chicago O’Hare International airports during one year. Results show that by the introduction of this technique it is possible to define larger scopes, partially reducing the amount of unrecovered delay.     

Finding time‐robust fuel‐efficient paths for a call‐taxi in a stochastic city road network

Intra‐city commuting is being revolutionized by call‐taxi services in many developing countries such as India. A customer requests a taxi via phone, and it arrives at the right time and at the right location for the pick‐up. This mode of intra‐city travel has become one of the most reliable and convenient modes of transportation for customers traveling for business and non‐business purposes. The increased number of vehicles on city roads and raising fuel costs has prompted a new type of transportation logistics problem of finding a fuel‐efficient and quickest path for a call‐taxi through a city road network, where the travel times are stochastic. The stochastic travel time of the road network is induced by obstacles such as the traffic signals and intersections. The delay and additional fuel consumption at each of these obstacles are calculated that are later imputed to the total travel time and fuel consumption of a path. A Monte‐Carlo simulation‐based approach is proposed to identify unique fuel‐efficient paths between two locations in a city road network where each obstacle has a delay distribution. A multi‐criteria score is then assigned to each unique path based on the probability that the path is fuel efficient, the average travel time of the path and the coefficient of variation of the travel times of the path. Copyright © 2016 John Wiley & Sons, Ltd.     

Real-time airport surface movement planning: Minimizing aircraft emissions

This paper presents a study towards the development of a real-time taxi movement planning system that seeks to optimize the timed taxiing routes of all aircraft on an airport surface, by minimizing the emissions that result from taxiing aircraft operations. To resolve this online planning problem, one of the most commonly employed operations research methods for large-scale problems has been successfully used, viz., mixed-integer linear programming (MILP). The MILP formulation implemented herein permits the planning system to update the total taxi planning every 15 s, allowing to respond to unforeseen disturbances in the traffic flow. Extensive numerical experiments involving a realistic (hub) airport environment bear out that an estimated environmental benefit of 1–3 percent per emission product can be obtained. This research effort clearly demonstrates that a surface movement planning system capable of minimizing the emissions in conjunction with the total taxiing time can be beneficial for airports that face dense surface traffic and stringent environmental requirements.     

Airport taxi planning: Lagrangian decomposition

The airport taxi planning (TP) module is a decision tool intended to guide airport surface management operations. TP is defined by a flow network optimization model that represents flight ground movements and improves aircraft taxiing routes and schedules during periods of aircraft congestion. TP is not intended to operate as a stand‐alone tool for airport operations management: on the contrary, it must be used in conjunction with existing departing and arriving traffic tools and overseen by the taxi planner of the airport, also known as the aircraft ground controller. TP must be flexible in order to accommodate changing inputs while maintaining consistent routes and schedules already delivered from past executions. Within this dynamic environment, the execution time of TP may not exceed a few minutes. Classic methods for solving binary multi‐commodity flow networks with side constraints are not efficient enough; therefore, a Lagrangian decomposition methodology has been adapted to solve it. We demonstrate TP Lagrangian decomposition using actual data from the Madrid‐Barajas Airport. Copyright © 2011 John Wiley & Sons, Ltd.     

A fuzzy approach to addressing uncertainty in Airport Ground Movement optimisation

Allocating efficient routes to taxiing aircraft, known as the Ground Movement problem, is increasingly important as air traffic levels continue to increase. If taxiways cannot be reliably traversed quickly, aircraft can miss valuable assigned slots at the runway or can waste fuel waiting for other aircraft to clear. Efficient algorithms for this problem have been proposed, but little work has considered the uncertainties inherent in the domain. This paper proposes an adaptive Mamdani fuzzy rule based system to estimate taxi times and their uncertainties. Furthermore, the existing Quickest Path Problem with Time Windows (QPPTW) algorithm is adapted to use fuzzy taxi time estimates. Experiments with simulated taxi movements at Manchester Airport, the third-busiest in the UK, show the new approach produces routes that are more robust, reducing delays due to uncertain taxi times by 10–20% over the original QPPTW.     

A two-stage taxi scheduling strategy at airports with multiple independent runways

Long taxiing times at large airports lead to fuel wastage and dissatisfied passengers. This paper investigates the 4D taxi scheduling problem in airports to minimize the taxiing time. We propose an iterative two-stage scheduling strategy. In the first stage, all aircrafts in a current schedule period are assigned initial 4D routes. In the second stage, landing aircrafts that are unavailable to fulfil their initially assigned routes are rescheduled using a shortest path algorithm based approach. In this paper, the simplified model used in most existing literature, that depicts a runway as having a single entrance and a single exit or even sets only one point to represent both of them has been discarded. Instead, we model the fact that a runway has multiple entrance and exit points and use an emerging concept—Runway Exit Availability (REA)—to measure the probability of clearing a runway from a specific exit during a specific time interval so that the taxiing scheduling model can be much higher approximation to the practical operation. An integer programming (IP) model factoring REA is proposed for assigning 4D taxiing routes in the first stage. The IP model covers most practical constraints faced in airport taxiing procedures, such as the rear-end/head-on conflict constraint, runway-crossing constraint, take-off/landing separation constraint, and taxi-out constraint. Besides, flight holding patterns at intersections are much more realistically modelled. Furthermore, to accelerate the solving process of the IP model, we have refined the formulation using several tricks. Simulation results by proposed scheduling approach for operations at the Beijing Capital International Airport (PEK) for an entire day demonstrate a surprising taxiing time saving against the empirical data and simulation results based on a strategy similar to what being used now days while showing an acceptable running time of our approach, which supports that our approach may help in real operation in the future.     

Fueling for contingencies: The hidden cost of unpredictability in the air transportation system

The aviation community is increasing its attention on the concept of predictability when conducting aviation service quality assessments. Reduced fuel consumption and the related cost is one of the various benefits that could be achieved through improved flight predictability. A lack of predictability may cause airline dispatchers to load more fuel onto aircraft before they depart; the flights would then in turn consume extra fuel just to carry excess fuel loaded. In this study, we employ a large dataset with flight-level fuel loading and consumption information from a major US airline. With these data, we estimate the relationship between the amount of loaded fuel and flight predictability performance using a statistical model. The impact of loaded fuel is translated into fuel consumption and, ultimately, fuel cost and environmental impact for US domestic operations. We find that a one-minute increase in the standard deviation of airborne time leads to a 0.88 min increase in loaded contingency fuel and 1.66 min in loaded contingency and alternate fuel. If there were no unpredictability in the aviation system, captured in our model by eliminating standard deviation in flight time, the reduction in the loaded fuel would between 6.12 and 11.28 min per flight. Given a range of fuel prices, this ultimately would translate into cost savings for US domestic airlines on the order of $120–$452 million per year.     

One manufacturer's approach to improving jet transport fuel efficiency

In the face of rapidly escalating costs of aircraft fuel, since the early 1970s, the aerospace industry has embarked on a program of measures to reduce fuel consumption. This article describes an integrated set of measures which have been undertaken by Lockheed‐California to improve the fuel requirements of the Tristar, a wide‐bodied three engine medium to long range aircraft. Measures include operational changes, improvements to maintenance practices, long term hardware modifications and technological advances.     

Effect of mass on multimodal fuel consumption in moving people and freight in the U.S.

The United States transportation sector consumes 5 billion barrels of petroleum annually to move people and freight around the country by car, truck, train, ship and aircraft, emitting significant greenhouse gases in the process. Making the transportation system more sustainable by reducing these emissions and increasing the efficiency of this multimodal system can be achieved through several vehicle-centric strategies. We focus here on one of these strategies – reducing vehicle mass – and on collecting and developing a set of physics-based expressions to describe the effect of vehicle mass reduction on fuel consumption across transportation modes in the U.S. These expressions allow analysts to estimate fuel savings resulting from vehicle mass reductions (termed fuel reduction value, FRV), across modes, without resorting to specialized software or extensive modeling efforts, and to evaluate greenhouse gas emission and cost implications of these fuel savings. We describe how FRV differs from fuel intensity (FI) and how to properly use both of these metrics, and we provide a method to adjust FI based on mass changes and FRV. Based on this work, we estimate that a 10% vehicle mass reduction (assuming constant payload mass) results in a 2% improvement in fuel consumption for trains and light, medium, and heavy trucks, 4% for buses, and 7% for aircraft. When a 10% vehicle mass reduction is offset by an increase in an equivalent mass of payload, fuel intensity (fuel used per unit mass of payload) increases from 6% to 23%, with the largest increase being for aircraft.     

Using on-board logging devices to study the longer-term impact of an eco-driving course

In this paper the long-term impact of an eco-driving training course is evaluated by monitoring driving behavior and fuel consumption for several months before and after the course. Cars were equipped with an on-board logging device that records the position and speed of the vehicle using GPS tracking as well as real time as electronic engine data extracted from the controller area network. The data includes mileage, number of revolutions per minute, position of the accelerator pedal, and instantaneous fuel consumption. It was gathered over a period of 10 months for 10 drivers during real-life conditions thus enabling an individual drive style analysis. The average fuel consumption four months after the course fell by 5.8%. Most drivers showed an immediate improvement in fuel consumption that was stable over time, but some tended to fall back into their original driving habits.