Integrated Airline Fleet and Crew Recovery through Local Search
Philip de Bruin, Marjan van den Akker, Kunal Kumar, Lisanne Heuseveldt, Marc Paelinck · 2025
Read Integrated Airline Fleet and Crew Recovery through Local Search by Philip de Bruin, Marjan van den Akker, Kunal Kumar, Lisanne Heuseveldt, Marc Paelinck (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-04274.txt.
Introduction
Philip de Bruin and four co-authors tackle airline disruption recovery with an integrated local search in this 2025 paper. Flight schedules are tight, so a single delay can ripple through aircraft and crews. Recovery must happen during the day, near real time, at low cost and low passenger harm. Most tools fix planes and crews separately. This team built a fast simulated annealing method that repairs both schedules together. Neighbor generation variants and recovery strategies were compared on real KLM data.
The integrated search resolved disruptions within about thirty seconds. Against naive delay propagation, it cut non performance costs by forty percent. It also beat the traditional split approaches. Speed matters as much as quality here: a perfect plan that arrives late is useless. The authors claim the first local search method for integrated recovery.
Catalog record
| CREATOR | Philip de Bruin, Marjan van den Akker, Kunal Kumar, Lisanne Heuseveldt, Marc Paelinck |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-04274
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Integrated Airline Fleet and Crew Recovery through Local Search Philip de Bruin, Marjan van den Akker, Kunal Kumar, Lisanne Heuseveldt, Marc Paelinck Abstract Airline operations are prone to delays and disruptions, since the schedules are generally tight and depend on a lot of resources. When disruptions occur, the flight schedule needs to be adjusted such that the operation can continue. Since this happens during the day of operations, this needs to be done as close to real time as possible, posing a challenge with respect to computation time. Moreover, to limit the impact of disruptions, we want a solution with minimal cost and passenger impact. Since airline operations include many interlinked decisions, an integrated approach leads to better overall solutions. We specifically look at resolving these disruptions in both the aircraft and crew schedules. Resolving these disruptions is complex, especially when it is done in an integrated way, i.e. including multiple different resources. To solve this problem in an integrated manner, we developed a fast simulated annealing approach. To the best of our knowledge, we are the first to develop a local search approach to resolve airline disruptions in an integrated way. This approach is compared with traditional approaches, and an experimental study is done to evaluate different neighbour generation methods, and to investigate different recovery scenarios and strategies. The comparison is done using real world data from KLM Royal Dutch Airlines. Here, we show that our approach resolves disruptions quickly and in a cost-efficient manner, and that it outperforms traditional approaches. Compared to naive delay propagation, our method saves 40% in non-performance costs. Moreover, while most airlines use tools that consider resources separately, our approach shows that integrated disruption management is possible within 30 seconds. \newpageafter abstract 1 Introduction Airline operations are prone to disruptions. These could be due to bad weather conditions, airport congestion, or other unforeseen circumstances. Airline schedules are generally tight and depend on a lot of resources, thus disruptions may have a big effect in the sense that they easily propagate through the schedule. Improper handling of disruptions can lead to many cancellations, which can cause even more chaos and large recovery costs. This creates additional operational challenges as it becomes harder to resume the original schedule. Disruptions also have a big impact on passengers, who need to be reaccomodated. In the EU, this translates to an estimated cost of 6.5 billion euros in 2024 ( Skycop, 2025 ) . On top of that, the negative impact on passenger satisfaction brings a competitive disadvantage for the airline. Overall, it is estimated that disruptions costs an airline around 8% of their total revenue ( CMAC, 2024 ) . Therefore, it is important to handle these disruptions correctly and efficiently, as they can become very costly. To make it worse, since the COVID pandemic, most airlines have been left with a lower availability of their resources, making airline operations even more vulnerable to disruptions. This is clearly visible in a recent performance review by Eurocontrol ( EuroControl, 2024 ) . They show that punctuality has been on a decline since the summer of 2021 and punctuality is also lower than 2019, while overall traffic levels are below 2019 levels. Examples of these resource issues are: shortages of flight crew and supply chain issues on aircraft parts, where the latter implies that aircraft stay in maintenance longer and are thus unavailable for flight. When a disruption occurs, an airline typically has a range of options to handle the situation. They can make changes to the resource assignment, essentially moving or swapping flights between, for example, aircraft. For example, when maintenance of an aircraft is delayed, it may not be able to perform the first flight planned after maintenance. In this situation, the airline may choose to reassign the flight to another aircraft, or swap assignments between two aircraft. Furthermore, they can make adjustments to the schedule, i.e. to delay or, in an extreme case, cancel the flight. These options each come with their own impact, including on passengers, schedule stability, and other key considerations, all which can be attributed to a cost. To avoid large losses, besides resolving a disruption, we also need to minimize these costs. Next to that, these disruptions happen live during the day of operation and may be revealed on a short notice. Consequently, they need to be resolved quickly to not ground the operation; and any changes made to the schedule need to be communicated to not only the passengers, but also other resource providers such as ground operations, airport, air traffic control, and others, as they need to adjust their schedules to incorporate the new adjustments. The need to communicate changes to all these resource providers adds a lot of time pressure for those involved in resolving these disruptions. For this reason, a schedule in which all disruptions are resolved needs to be found within at most a few minutes of computation time. This is known as the Airline Recovery Problem and is usually divided into aircraft, crew, and passenger recovery, each focussing on their respective part of the process. Since flight delays and cancellations are changes to the flight schedule, the inclusion of those actions is often referred to as schedule recovery. To solve the Airline Recovery Problem , one can consider separate recovery problems for each of the resources and solve them sequentially. Here, each part boils down to a hard optimization problem with its own computational requirements. However, these problems have a lot of interlinked decisions, meaning that decisions in one stage may lead to conflicts or infeasibilities in the other. For example, the delay of a flight might be preferable for the aircraft assignment, but may mean that the assigned crew exceeds their maximum allowed duty hours. Considering an integrated approach mitigates these issues, resulting in better overall solutions. However, integrating decision-making with multiple resources into one model adds more complexity, increasing the model size considerably, which makes it more challenging to find a good solution in a short amount of time. Although many papers focus on a single resource, the integrated approach has been studied before by, for example, Petersen et al. (2012) , and Arikan et al. (2017) . Here, the disruptions under consideration are mainly related to the reduction of airport capacity. There has not been much focus on the unexpected unavailability of aircraft or crew members. These studies also show models with a typical runtime up to 15 to 20 minutes, meaning that application in practice is limited. Lastly, in most of the earlier work, only a single aircraft type is considered, meaning that all crew can fly on all the aeroplanes available. However, when swapping aircraft of different types, we have to deal with complications due to crew not being licensed to fly a certain aircraft type, increasing the complexity of the problem. In this paper, we present a local search algorithm that solves the Airline Recovery Problem in an integrated manner and as close to real-time as possible. The latter gives an airline enough time to implement the suggested changes, making the algorithm fast enough for application in practice. We focus on aircraft and crew schedules, but we also take the passenger