Agent-Based Modelling and Simulation, airline disruption management, airline operations, Airline Operations Control, Coordination, Decision Support Systems

A novel approach to developing decision support systems for airline disruption management

Airlines continuously face disturbances that disrupt their highly optimized schedule. Events like adverse weather conditions, air traffic management decisions, sick crewmembers, or aircraft mechanical problems often cause delays in the airline’s schedule. For an airline such disruptions are very costly because they tend to cause domino effects in the air transportation schedule. Over the year 2007 alone, U.S. carriers lost over $8  billion because of delays of some sort Barnhart (2009). Therefore, reducing the impact of disruptions on the airline schedule could considerably reduce these costs. An airline usually recovers from such disturbances through the role played by the Airline Operations Control (AOC) centre. If a disruption affects flight plans, then human operators at the AOC center take corrective actions in real-time in order to manage the disruption. Possible actions include cancelling or delaying of flights and swapping aircraft or crew, and are often the result of decision-making and coordination processes that involves many AOC operators.

The specific organization of an AOC center depends on multiple factors. These factors include the airline size, type of airline operations, location, and airline culture. However, despite the different organization types, it is possible to identify operators that are common to AOC centers. Fig. 1 gives an overview of a typical AOC center showing the operators, the technical systems, and the interactions between them and their external world (while the exact terminologies may vary per airline). It should be noted that in addition to the operators shown in Fig. 1, there exist other services in AOC centers which provide support for AOC operators (e.g. operational engineering). In addition, a crisis center which coordinates activities after an accident or incident is often an integrated part of an airline’s AOC center.


        Fig. 1 AOC embedded in the large air transportation system

To date, studies to improve AOC performance have mainly focused on using optimization techniques to develop decision support tools. For instance, Bratu & Barnhart (2006) propose two optimization tools that generate recovery plans for aircraft, crews, and passengers by determining which flight leg departures to postpone and which to cancel. Abdelghany et al. (2008) propose a decision-support tool that provides AOC centres with the capability to develop a proactive schedule recovery plan that integrates all flight resources. The optimization tool examines possible resource swapping and flight requoting to generate a schedule recovery that minimizes flight delays and cancellations. Petersen et al. (2012) propose a mixed-integer programming tool to solve the fully integrated airline recovery problem including the schedule, aircraft, crew, and passenger problems. In the same vein, Arikan et al. (2017) propose an optimization tool to solve the fully integrated airline recovery problem using a conic quadratic mixed integer programming formulation. Santos et al. (2017), present an integer linear programming tool to help AOC controllers decide which flights to delay and which flights to make depart on time.

From a combinatorial optimization perspective, these tools have the mathematical capability in minimizing airline operating costs and passenger costs. However, such a combinatorial optimization approach fails in capturing the complex socio-technical nature of AOC (Feigh & Pritchett, 2010; Bruce, 2011; Richters et al. 2017). According to Clausen et al. (2005), there is a gap between the support offered by IT systems and the reality faced in AOC centres. In order to address these socio-technical challenges, Delft Aviation proposes a novel approach for developing and validating decision support tools that makes it possible to learn in the early design phase from simulated disturbances, and to feedback this learning to the further improvement of the early design. Through the iterative cycle of design improvement and assessment, shown in figure 2., the resulting overall socio-technical design will become resilient against a wide range of disturbances. Agent-based modelling and simulation is chosen because it has been extensively used to model and analyse complex socio-technical systems, and address cases where agents need to coordinate and solve problems in a distributed fashion.


Fig. 2 AOC performance assessment feedback to decision-support systems designers

We can help you improve the performance of your airline operations with one of our services below:

  • Evaluate from the complex socio-technical perspective the decision-support systems on their performance and practical implementation. The evaluation is conducted using both a scenario-based and task-based analysis and aims at providing feedback to improve tool design. E.g. which activities should automation aim to support? And how should the work be split between the decision support system and human operator?
  • Evaluate your current airline disruption management policies using agent-based modelling and simulation.
  • Propose innovative disruption management policies for a wide range of scenarios based on recent advances in psychology.
  • Organize training workshops for airline controllers (especially novices) to improve their decision making and coordination processes.


  • Abdelghany, K.F., Abdelghany, A.F., Ekollu, G. (2008). An integrated decision support tool for airlines schedule recovery during irregular operations. Eur. J. Oper. Res. 185(2):825-848
  • Arikan, U., Gurel, S., Akturk, M.S. (2017). Flight network-based approach for integrated airline recovery with cruise speed control. Transportation science. Articles in advance, pp. 1-29.
  • Barnhart, C. (2009). The Global Airline Industry, chapter Irregular Operations, pages 253–274. John Wiley & Sons, Ltd, West Sussex.
  • Bratu, S., Barnhart, C. (2006). Flight operations recovery: New approaches considering passenger recovery. J. Scheduling 9(3):279-298
  • Bruce, P. J. (2011). Decision-Making in Airline Operations: The Importance of Identifying Decision Considerations. International Journal of Aviation Management, 1(1/2):89–104.
  • Clausen, J., Larsen, A., Larsen, J., 2005. Disruption Management in the Airline Industry – Concepts, Models and Methods.
  • Feigh, K. M. (2008). Design of Cognitive Work Support Systems for Airline Operations. PhD thesis, Georgia Institute of Technology.
  • Petersen et al. (2012) propose a mixed-integer programming tool to solve the fully integrated airline recovery problem including the schedule, aircraft, crew, and passenger problems. Richters, F., Schraagen, J.M., Heerkens, H. (2017). Assessing the structure of non-routing decision processes in Airline Operations Control. Ergonomics, 59:3, 380-392.
  • Santos, B.F., Wormer, M.E.C., Achola, T.A.O., Curran, R. (2017). Airline delay management problem with airport capacity constraints and priority decisions. Journal of Air Transport Management 63. 34-44.


The cover photo in this blog post shows KLM’s AOC centre. Photo credit: