Aircraft boarding methods
Project name: Aircraft boarding methods
Class: 4IT496 Simulation of Systems (WS 2014/2015)
Author: Dinara Mansurova
Model type: Agent-based simulation
Software used: NetLogo
This simulation explores three aircraft boarding methods: Random, WMA and Back-front WMA.
Airports serve thousands of people every day and one of the main problems they meet is the efficient passenger boarding. In order to increase productivity and minimize the time needed for boarding several methods were created.
- random (which is widely used, the most common): people just taking their seats without any specific order
- WMA (Window-Middle-Aisle) - boarding is divided into three groups: seats near the windows, middle seats and aisle seats. The first boarding group is windows then middle and finally the aisle group.
- Back-front WMA: we divide groups in the same way as in WMA, but the only difference, that we sort passengers in decreasing order (one group is two window seats, 2 middle seats and 2 aisles, in short it can be described as by row boarding) . For each group boadring starts from back of the plane.
To simulate different boarding scenarious and try to find out the optimal one.
Environment: NetLogo 5.2.1
NetLogo was chosen as a working environment due to the nature of the problem. Although the problem answers for finding the optimal way, it requires ulti-agent modeling enviroment, as the communication between agents takes place.
Agents (or turtles in NetLogo vocabulary) are plane passengers. Passenger agent has three attributes: gotoX, gotoY and luggage-status. In short, they contain information about seat location and luggage.
By default, plane is always full. We have only one type of agents, but we differenciate passengers with luggage and those who travel light. For the firts case, usual "person" shape was chosen. Passengers with luggage use "person business" shape. At the beggining of the simualation all agents are green. When passenger occupies his seat, the colors is chaged into red
Detailed description of the simulation
In the simulation boarding method, percent of passengers with luggage can be chosen. Other parameters, as luggage-load-time, change-place-time, number-of-passengers, are fixed. Plane has 18 rows and 3 seats per each side, in total 108 seats (number of seats = number of passengers).
The grafical user interface of the model is shown below.
The plane map is very simplified in order to be more understandable. On the left side of the interface three buttons, one slider, one chooser and monitor can be found.
"Setup" button creates the plane map and passengers.
"Go" and "Go (always)" start the simulation. The difference is in the way they behave, the first one shows one step per click, when the second is responsible for the full simulation.
Percentage of passengers with luggage can be chosen.
The preffered boarding is chosen.
Monitor displays the total number of time needed for boarding. It counts the ticks and agent delays (luggage load or time needed for place change). One delay is one tick.
How does the simulation work
1. Firts of all the "world" is created with the "Setup" button.
2. The boarding type and Luggage percentage is chosen.
3. Then the simulation is simply started with the "go" button.
If the passenger occupies his seat he turns red. The passengers with luggage after the loading baggage change their shape into usual "person" and the loaded bag is depicted as a pink patch on the shelfes. Agents are moving one by one (one agent on one patch).
Simulation is over when all passengers are on their seats.
I run the simulation 10 timer per each type of boarding and with different values of the luggage percentage parameter. On the table below results of the simulation are shown.
It is obvious from the simulation result table that WMA method is the fastest one. If we compare results fro different luggage parameters it is still leading the pack. On average WMA requres 109 ticks for boarding.
On the other hand, front-back WMA, which was also expected to be fast, is loosing ground to WMA with the average 164 number of ticks. It is quite fast, but due to delays it is not the fastest one.
Random boarding is 230 ticks on average. If we will take a look at the dependencies between luggage percentage and number of ticks for different methods, we can conclude that number of ticks grows for greater amount of luggage (due to delays of course), but still the difference between methods remains almost the same. It means, that luggage is not affecting the boarding method itself.
Effective boarding can save costs of the airline companies. As we can see, there is a considerable difference between Random seat allocation and WMA. This simulation project showed that WMA boarding is the most efficient boarding method.
On the contrary, all special boarding methods have special rules and requiremens for passenger order and seat allocation, that's why it is difficult to apply them in real life. Additionaly, real agents, humans behave very unpredictably, as well as their environment.
In this section simulation code can be found. The code is in the NetLogo model file with .nlogo suffix.
The simulation itself is available in a form of a zip archive.