1 CASE STUDY 7: THE BUS STATION

1.1 Introduction
In the past, a bus station would be made up of a large empty space where regional and city
buses each had their own area. One or several bus lines always used the same platforms
where they could let passengers get on and off.
Thanks to the progress of technology, passengers can now see the arrival time and location
for their buses on matrix display boards. As a result, the so-called rigid distribution (bus
line 5 always arrives at platform 3) can be broken (bus line 5 arrives in 5 minutes at platform
3). With this type of technology, bus line 5 might arrive sometimes at platform 3 and other
times at platform 5. Further, it is now thought unnecessary and even annoying to have buses
occupying a platform space when they are not due to leave for some time. By assigning a
separate alighting platform, from where the bus drives to a parking area (buffer), only mov-
ing to the departure platform a couple of minutes prior to departure. These two pillars form
the basis of the Compact Dynamic Bus station concept: dynamic distribution and separation
of processes.

According to logistics concepts for a better utilization of capacities, it is desirable to attempt

to the same process with fewer platforms. This is useful in where space is limited, expensive
or required for other purposes. The core question for any town or city wishing to implement
such a new bus station is: how many platforms are required in the new situation and how
large does the buffer need to be?
The answer depends strongly on the timetables and since a new bus station is a big invest-
ment, a simulation study is a very wise thing to do.

1.2 Situation description

A city with 150.000 inhabitants has an outdated bus station and too little space. The City
Council gives an order for a simulation study regarding a compact dynamic bus station. The
aim of the study is to determine the number of platforms required and the size of the buffer.
The study is carried out for a typical period: the Monday morning rush hour.

All buses will go through the bus station, thus city buses, regional buses and special lines,
such as school buses. In total, there are 37 lines, and each line is assigned a preferred plat-
form. The bus guidebook is described in the Excel file Bus.xls. An explanation of the contents
and structure of this file including 275 buses during the typical period is given in annex 1.

Buses arriving more than 5 minutes before the departure time mentioned in the schedule let
the passengers get off at a remote alighting location, which takes 30 seconds to complete.
Afterwards, the buses drive to the buffer and wait there until 2 minutes before departure
time. The buses then drive to the station, the area with the platforms, bus stops and passen-
gers.
Buses arriving less than 5 minutes before the departure time drive on to the platforms di-
rectly, where the alighting of passengers and the boarding of new ones occur simultane-
ously. This process takes 2 minutes, but buses do not leave before the time mentioned in the
schedule.

As a result, buses combining boarding and alighting remain at the most 5 minutes at the sta-
tion and other buses only 2 minutes. The time necessary to drive from the alighting location
to the buffer and from the buffer to the station were kept outside this model, as well as the

Case study 7 The bus station ©Incontrol Enterprise Dynamics 1

organisation of the available space and the passengers’ behaviour. A layout of the process is
shown in Figure 1-1.

Bus stop 1b Bus stop 1a

Bus stop 2b Bus stop 2a

Descend 1
Bus stop nb Bus stop na
Descend 2 Buffer

Arrival > 5 min before Depart

Descend n Arrival <= 5 min before Depart

Figure 1-1: Bus station layout

Each platform consists of two bus stops; so two buses can take position behind each other.
The rear bus can only leave when the front stop is empty. Therefore, when the platform is
empty, the first bus to arrive will take the front position. When the rear stop is occupied, the
front stop cannot be reached. A bus on the rear stop does not move to the front stop if it sub-
sequently becomes vacant.

Second Stop

First Stop

Figuur 1-2: Order of bus stops on the platform

For the passengers’ convenience, the bus lines have a preferred platform (see bus.xls). If this
platform is occupied, only the nearest platforms are used. So a bus with preferred platform 4
is only allowed to divert to platform 3 or 5. As far as the first and last platforms are con-
cerned, only the single adjacent platform is used.

In this basic model, all time operations are temporarily constant (deterministic), which
means that there are no disturbances as a result of, for instance, buses arriving too early or
too late, or from longer or shorter boarding times. By making this model deterministic, it is
easier to validate it. At a later stage, possible disturbances can be added easily.

Case study 7 The bus station ©Incontrol Enterprise Dynamics 2

1.3 Assignments

1. Is it possible to make a calculation beforehand or a reasoned estimate of the number of

platforms required?

2. Create a simulation model with 6 platforms based on the preferred distribution in bus.xls
and determine the required size of the alighting location and of the buffer.
N.B.: You can do so by reading bus.xls directly from Excel or by cutting and pasting
from bus.xls into an Arrival list (busstation1.mod and busstation2.mod).

3. Add a disturbance with a one-minute standard deviation to the deterministic arrival pat-
tern and investigate to what extent the results of the deterministic model in the previous
scenario are still valid.

*4. Make your own distribution of preferred platforms and investigate if it is possible to
achieve comparable results with fewer platforms.

Case study 7 The bus station ©Incontrol Enterprise Dynamics 3

ANNEX 1: THE BUS GUIDEBOOK

The bus guidebook bus.xls is going to be developed in this annex. In the guidebook, you can
find the schedule of 275 buses on 37 lines during a normative period. The starting time 0
corresponds to the beginning of the schedule on Monday morning and the period studied
covers about 4 hours.

Bus.xls is divided into 3 parts: the tab ‘Bus schedule’ containing all 275 buses sorted accord-
ing to their arrival time, the tab ‘Platform-view’ and finally the 37 bus lines.

1. Bus schedule
The figure below displays the first 10 buses of the ‘Bus schedule’ tab. This tab features all
buses arriving at the bus station during the rush hour. Each row stands for a bus.

Figure 1-1: bus.xls, ‘bus schedule’ tab, which contains the whole schedule

Column A:
Arrival Time: The time of arrival into the model (in seconds). According to the schedule,
this is 5 minutes before arrival (see column E).

Column B:
Atom Name: The number of the bus line.

Column C:
Quantity: The number of buses, which is entered in order to paste the data into an arrival
list easily. The format of this sheet is namely the same as for the arrival list.

Column D:
Channel: The channel used to send out the bus. It is also entered in order to paste the data
into an arrival list easily.

Column E:
Arrival: The arrival time according to the schedule (in seconds), which is a different time
than the arrival time into the model. The purpose of this distinction is to allow the addition
of a disturbance into the model, causing buses to arrive at the bus station earlier or later than
the arrival time.

Column F:
Depart: The departure time (in seconds), according to the schedule.

Case study 7 The bus station ©Incontrol Enterprise Dynamics 4

Column G:
Platform: The preferred platform.

Column H:
Inter Arrival Time: Inter-arrival time in the model (in seconds), which is the time elapsing
between the preceding bus and the current bus. This column has been added for the reading
of these times from Excel.

2. Platform View
The ‘Platform-view’ tab (Figure 1-2) gives an overview of the bus lines and their corre-
sponding preferred platforms. The purpose of this worksheet is only to give an overview and
may not be adjusted. In order to change the preferred platform of a bus line, the tab of the
relating bus line has to be opened (Figure 1-3).

Figure 1-2: bus.xls, ‘platform-view’ tab; overview of the platform distribution

3. The bus lines

The Figure below is an overview of the schedule of bus line 1, to be found under tab ‘1’.
The layout of the columns corresponds to the layout of columns A to G in the ‘Bus sched-
ule’ tab (see Figure 1-1). In order to alter the preferred platform, the red figure behind prior-
ity platform has to be changed into the number of the preferred platform. The preferred plat-
form is automatically adjusted in the tabs ‘bus schedule’ and ‘platform-view’ as well.

Figure 1-3: bus.xls, tab ‘1’; schedule for bus line 1

Case study 7 The bus station ©Incontrol Enterprise Dynamics 5
2 SUGGESTIONS FOR MODELLING IN ED

Buses are entering the system either from an arrival list that you can obtain by cutting and
pasting from bus.xls, or from a Source, which is directly linked to bus.xls. In both cases, extra
labels have to be made to denominate the data that is to be found in the columns Arrival,
Depart and Platform of the Excel file.

After they have entered the system according to the arrival time mentioned in bus.xls, the
buses first have to wait a constant 5 minutes. Alternatively, to examine a stochastic arrival
pattern, choose a normal distribution with an average of 5 minutes and a standard deviation
of one minute. For more details, see annex 1.

Multiservice atoms can be used to handle several buses at the same time. It is possible to
give all buses the same cycle time or to give each bus a different cycle time. Use these atoms
for holding buses for 5 minutes and also for the alighting process and the buffer.

Use one queue as a distribution tool for the various platforms. The platforms themselves can
be modelled with Servers and a supply buffer.

The input or output of an atom can be closed or opened by using the functions closeinput,
closeoutput, openinput and openoutput. By using an atom reference, this can also be applied
to another atom, for instance closeinput(out(1,c)).

In a model linked to Excel, the T082-Excel atom has to be dragged into the model. Read the
Help file for more details. In this atom, a link can be made with an Excel file. Afterwards,
the function excelread(a,b) can be used, in which a symbolizes the row and b the column.

When all the buses are created, there is no inter-arrival time in Excel any more. ED will read
this as 0 and will then go on creating new atoms. Therefore, after 275 buses have entered the
system, you can stop the creation of new atoms by using the closeoutput(c) function.

Case study 7 The bus station ©Incontrol Enterprise Dynamics 6