- Group Name: PlaneBoarding
- Group participants names: Nils Blach, Anton Schäfer
- Project Title: Airplane Boarding
- Programming language: Python
Reducing airplane turn around time is of interest to all airlines, as it leads to more effiencent use of airport and airline resources, yielding great economical benefits. Still the major portion of turn around time, the boarding process, is oftentimes unnecessarily inefficient. Using our model, we aim to identify causes of inefficiencies and strategies to overcome them.
The model is agent-based an fcouses on passengers' behavious in the aisle. In each discrete time step of 0.1 s, all actors in the aisle can act. The actors behaviour can be described with a finite state machine. Actors start in state 0, where they are outside the plane, then enter the plane and search for space to store their luggage. After storing their luggage they move to their seat and sit down. As soon as all passengers are seated, boarding is completed and the simulation ends.
We also included an animation of the simulation. To use it, run a simultation, and generate an Animation a = Animate(sim) if sim is a simulation object. Then, call a.animate() press the space bar to start the animation, and watch the passengers board the plane. You can also just execute the file main.py to see the Steffen method with the Bombardier CS100 (You will need pygame).
How does our model compare to the one from Van Landeghem and Beuselinck? How do the Steffen method and other boarding methods compare to each other, when applied for different planes? How does luggage load influence boarding time?
We expect our model to yield similar results as the one given by Van Landeghem and Beuselinck. The Steffen method is supposed to be the most effective boarding method, so we expect it to perform best. We expect that the boarding time drastically increases with increasing luggage load.
Van Landeghem, H, and A Beuselinck. "Reducing Passenger Boarding Time in Airplanes: A Simulation Based Approach." European Journal of Operational Research, vol. 142, no. 2, 2002, pp. 294–308.
Steffen, Jason H. ''Optimal boarding method for airline passengers." Journal of Air Transport Management 14.3 (2008): 146-150.
Data Passenger Size from: NASA. “Man-Systems Integration Standards.” Man-Systems Integration Standards, vol. 1, July 1995, p. 30.
Barrett, Sean D. ''How do the demands for airport services differ between full-service carriers and low-cost carriers?." Journal of air transport management 10.1 (2004): 33-39.
Images in Animation:
Seat: www.conceptdraw.com/solution-park/building-seating-plans (accessed 29.11.2018)
Plane nose and tail: https://www.google.ch/url?sa=i&source=images&cd=&ved=2ahUKEwiqrpCC0fneAhVRDOwKHRvzDH0QjRx6BAgBEAU&url=http%3A%2F%2Fcvfreeletters.brandforesight.co%2Felegant-frontier-airlines-planes-seating-chart%2Felegant-frontier-airlines-planes-seating-chart-15-elegant-united-seating-chart-masterlistforeignluxury&psig=AOvVaw0mpEvRKrlOSIJLULGzb9s5&ust=1543581837081380 (accessed 29.11.2018)
We use an agent-based model.
Before reproducing our work, make sure you have pyhton3, numpy, and matplotlib installed. If you want to see an Animation of the Steffen method with the Bombardier CS100, you will need pygame. To run the light test, just run the python script light_test.py. It will simulate boarding of an Airbus A320-200 with 100 % load and 180 passengers, using the random seat assignment and the Steffen method. It will print the boarding times to the console. If you want to see the animation mentioned above (not part of the reproducability test) then execute the main method and press space after the window with the graphics appears. To run the full test, first run the python script full_test.py to generate the data. After it produced all the CSV files, just run the pyhton script full_test_graphics.py to produce all graphs used in the paper.