Insights into the potential of robust capacity planning at KLM Aircraft Services

Harmsen, J.A.J. (2012) Insights into the potential of robust capacity planning at KLM Aircraft Services.

Abstract:This report is written for KLM Aircraft Services, the department of KLM which is responsible for moving and preparing aircraft for flight at the platforms, and gives insight into the possibilities of creating and the potential of using a more robust capacity planning. Using the current deterministic way of planning, it is not possible for Aircraft Services to create an explicit link between covering of workload (the number of operators needed to perform the amount of work which needs to be done at a certain moment in time) and on-time performance. Resulting from this, the planning gives just one option for the workload, independent on the variability in the arrival punctuality and/or process duration. When we take uncertainty and unforeseen events into account in the planning methods of Aircraft Services, we take deviations from the fixed schedules and timetables into consideration and cope with possible disturbances. This leads to a robust way of capacity planning. A robust capacity planning takes possible delays, which can occur at an operational level, into account in the planning phase. For this reason, it works well in many situations, instead of perfect in just one situation. This project has the following goal: “To provide insight into the effect of uncertainty and unforeseen events on the dynamics of personnel capacity planning, related to the performance of KLM Aircraft Services.” To achieve the goal of this project, it is only necessary to analyse one of the processes of Aircraft Services in depth. The insights of this analysis are then extended to the other processes. To choose between the different processes of Aircraft Services, we first characterised them and afterwards made an assessment, where we decided to choose for the refuelling service. The most important characteristics to characterise the services of Aircraft Services in the scope of this research are:  The type of process: strict arrival/departure (executed exactly at arrival or at departure) or arrival/departure oriented (executed after arrival and before departure).  The flexibility of scheduling the process (dependency on other processes and whether the process has flexible permissible time windows).  The duration of the process.  The variation in the duration of the process. We analyse the different elements affected by uncertainty and unforeseen events in the planning of KLM. A part of the variability in these elements is relatively predictable and a part is truly uncertain. The most important elements in the planning are the arrival punctuality of aircraft, the duration of the ground processes and possible disturbances. To describe the behaviour of these elements, we use theoretical probability distributions. We create four models to get from a predefined flight schedule to the corresponding performance of the refuelling process, while incorporating uncertainty and unforeseen events in the planning. Our first model uses Monte Carlo Simulation to generate workload lines based on the combination of predefined theoretical probability distributions and the flight schedule. We optimise the starting time per task using our second model: an optimisation heuristic. Our third model uses Linear Programming to fit a set of shifts to the optimised workload, on different levels of workload coverage. We determine the corresponding performance by using our fourth model: a Discrete Event Simulation, which we use to simulate a day of operation of the refuelling process. When we create a workload planning incorporating uncertainty and unforeseen events, we obtain a wide range of possible values the workload can take. From this, we conclude that uncertainty and unforeseen events have a big influence on the planned capacity. The eight hour shifts used for scheduling have a large impact on the robustness of the planning, since they cover, due to their duration, more than just the peaks in the workload, which leads to a lot of excess capacity. This allows peaks to move within a shift without large consequences for performance. The shifts thereby neutralize a large part of the discrepancy between planning and reality, caused by the arrival punctuality of flights. Simulating a day of operation results in high performances, but due to the assumptions made to be able to model the refueling process with the limited scheduling guidelines KLM could supply, our simulation model overestimates the actual performance. We cannot extent the simulated performance directly to reality, but we are able to extent the insights about dealing with uncertainty and unforeseen events in the workload planning to the other services of Aircraft Services. This is possible for the Refuelling service, Water service, Toilet service, Aircraft Handling Support and the Pushback and Towing service. For strict arrival/departure processes the variation in the arrival/departure punctuality of aircraft is most relevant and for arrival/departure oriented processes the variation in the task process time is most relevant. To deal with uncertainty, we recommend implementing an approach where Aircraft Services manages the current amount of uncertainty and simultaneously start initiatives to reduce this uncertainty. Finally Aircraft Services needs to buffer against the amount of uncertainty which cannot be reduced. To manage uncertainty, Aircraft Services needs to make a good forecast for the parameters of the defined probability distributions. These probability distributions have to be used in the process of creating workload profiles (instead of the current planning norms for tasks) to create an estimate of the range of values the workload can possibly take. Then KLM needs an integrated planning approach to make well founded decisions about which parts of the workload to cover, which shifts to use and how to schedule other activities (like breaks). This integrated planning approach is a process with multiple cycles of design, testing and adapting, which has to be executed by an expert with a lot of knowledge of all the processes. Reducing uncertainty starts with executing further research to define the impact of all sources of uncertainty and afterwards taking steps to decrease the most significant sources. We recommend reducing uncertainty in an iterative way, alongside managing and buffering against uncertainty. KLM has to buffer against the amount of uncertainty and unforeseen events which cannot be reduced. Buffering against uncertainty is according to the Law of Buffering automatically being done by some combination of inventory (or cycle time), capacity and time. Arrival/departure oriented processes (Refuelling, Water, Toilet) can buffer with all three options. For strictly arrival/departure processes (Pushback, AHS) buffering with time is not desirable. KLM has to remember that introducing buffers in the planning always increases the robustness; however this goes at the expense of optimality.
Item Type:Essay (Master)
KLM Aircraft Services, Amsterdam, the Netherlands
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:85 business administration, organizational science
Programme:Industrial Engineering and Management MSc (60029)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page