Evaluation of a Multi-Agent System approach to airline disruption management

Abstract

Each day, airlines face disturbances that disrupt their carefully planned operations. Events like adverse weather conditions, sick crew members, or damaged aircraft often result in delays in the airline's schedule. An airline recovers from such disruptions through the role played by its Airline Operations Control (AOC). A Multi-Agent System (MAS) approach to airline disruption management was recently proposed under the acronym MASDIMA (Multi-Agent System for Disruption Management in AOC). The purpose of this paper is to evaluate this MAS supported AOC approach on its performance and its practical introduction. This is done using a scenario-based analysis to compare the MAS supported policy to human-team based AOC policies. A task-based analysis identifies how well AOC is able to cover a set of tasks using the MAS supported policy. The scenariobased analysis shows that the MAS supported AOC is able to find the optimal solution, and to do this significantly faster. The taskbased analysis identified two main challenges for implementing the MAS supported AOC policy: i) to overcome the loss of experience that is caused by significantly automating humans roles in AOC, and ii) to reduce the workload for people that remain in AOC after its introduction. The paper concludes that implementing the MAS supported AOC policy leads to both better and faster resolutions, though the replacement of human roles also poses novel challenges that remain to be resolved: a potential increase in workload for the remaining human role and loss of experience in handling exceptional situations.

Formal modelling and verification of a multi-agent negotiation approach for airline operations control

“…In a previous work by the authors [43], the same scenario was evaluated for a different multi-agent system called MASDIMA [44]. The details of this evaluation are described in [45].…”

Agent-based Modelling and Simulation of Airline Operations Control Decision-Making under Uncertainty

“…The [1] namely an aircraft mechanical problem and a passenger connection problem. Previously, the aircraft mechanical problem scenario was considered in Bouarfa et al [7,8]. In this study, we increase the complexity of the scenario by combining two disruptions.…”

“…For the identification of the relevant AOC agents we follow [7,8]. The relevant agents are the Operations Controller (OC), Aircraft Controller (AC), Crew Controller (CC), Stations Operations Controller (SC) and Flight Dispatch (FD) who share the same common goal namely recovering from the disruption through collaborating with each other.…”

“…Modelling and simulating these decision-making processes is expected to yield novel insights into the effect of robust scheduling and operational uncertainties on disruption management. Earlier ABMS studies of AOC operations have been conducted by Bouarfa et al [7,8]. In [7] ABMS is used to evaluate and compare the socio-technical and socio-economic effects of four AOC coordination policies.…”

“…The results of [7] provide novel insights on the operational effects of each AOC coordination policy. In [8] ABMS is used to evaluate socio-technical and socio-economic effects of a Multi-Agent System (MAS) that was designed to automate key roles in the AOC center. The findings indicate that implementing a MAS supported AOC policy leads to both better and faster resolutions, though the replacement of human roles also poses novel challenges that remain to be resolved.…”