logo-print

Section Information Systems (IS)

2IMI30: Business Process Simulation

Lecturer: dr.ir. B.F. van Dongen
Date: 7-February-2017 until 7-April-2017
Time and Location: Tuesdays: 10:45 - 12:30 : Metaforum 8
Fridays: 15:45 - 17:30 : Flux 1.03
ECTS: 5

Course contents

Organizations are constantly trying to improve the way their businesses perform. To this end, managers have to take decisions on changes to the operational processes. However, these changes are never without consequences and often high costs are involved. Therefore, it is of the utmost importance that these decisions are supported by a thorough analysis of all possible consequences on the organization.

To gain insights into the consequences of decisions on an operational process, one often resorts to simulation studies. In these studies, simulation models are made of the operational process under consideration, taking into account the necessary elements, such as costs, resources and activities. These simulation models are then executed with different parameters, to gain insights into the consequences of different decisions on the basis of which a final decision is made.

It is clear that the construction of simulation models of an operational process is a far from trivial task. Deciding which elements of the operational process to take into account and which not is key to getting useful simulation results.

In this course, we use a discrete event simulation tool called Arena to execute simulations. This tool allows for the graphical definition of a simulation model, together with complex definitions of queue types, resource availability and so on.

Objectives

The focus of the business process simulation course is on the construction of simulation models, the execution of these models and the interpretation of the simulation results. After the course, students:

  • have a thorough understanding of the workings of a discrete event simulation,
  • understand (pseudo) random number generation,
  • understand distribution fitting and are capable of analyzing input data for use in simulation experiments,
  • are capable of performing “sanity checks” on their models,
  • can translate a description of an operational process to a simulation model with the purpose of solving a given problem,
  • can interpret simulation results, compare simulations of the same system with different parameters and recommend a solution to a given problem,

Prerequisites

It is advised (not compulsory) for students to have some background in Business Process Modelling, Probability Theory and Statistics, for example through the following courses:

  • 2II05 - Business Information Systems
  • 1BB30 - Business Process Modelling
  • 2DD20 - Pre-master Statistics
  • 2DI25 - Probability Theory

Lectures

The lectures will be given in English, twice a week, in two hour blocks.

Instructions

No instructions are scheduled for the course.

Examination

The exam consists of a group assignment, combined with an individual assignment. The assignment will be a description of an operational process, combined with a problem statement that has to be solved using simulation studies. A report is expected detailing which solution is proposed and why.

Furthermore, a follow-up question has to be answered individually. For this, no simulations have to be performed for this, but insights into the relevant research topics are required.

Finally, students have to grade themselves and their group members using a peer-assessment instrument.

OASE and Arena

To support this course both OASE and Arena are being used. All participants of this course are supposed to register via OASE. Here one can find announcements, slides, assignments, etc.

It is mandatory to use Arena for this course. Note that Arena requires Microsoft Windows. Linux and Mac OS are not natively supported and no support is provided for running Arena in a virtual machine.

Both the FactoryTalk Client Activation Tool and Arena 15 are provided by the central IT department on the following pages in English. If you're not using Internet Explorer, the installation files can be accessed through \\campusmp.campus.tue.nl\software\rockwell\arena\15.00_sp1. Note that these URLs can only be reached from the TU/e campus, it is not reachable from outside the campus. Detailed installation instructions are included on that page, as well as Arena Install Guide . Follow these instructions to install both Arena 15 and the FactoryTalk Activation Client.

To use Arena, 50 licenses are available to the students of this course. To get access to these 50 licenses, students have to provide their computer names.

Material

The material used in this course consists of the following book:

Title: Simulation Modeling and Analysis with Arena
Authors: Tayfur Altiok, Benjamin Melamed.
Publisher: Academic Press, © 2007.
ISBN 13: 978-0-12-370523-5
ISBN 10: 0-12-370523-1

The book includes version 11 of the Arena software. However, version 15 is used in the course. Other versions than 15 should not be used in the course, as we only provide licenses for version 15 and any unlicensed version is not sufficient.

Besides the book, slides and papers will be used as lecture material. The assignments, slides and other lecture material will be distributed via Canvas.

Assignment

The exact content of the assignments is to be determined. However, the assignments will be structured as follows:

  1. First, a design has to be proposed in writing. This design follows steps 1 and 2 of the model building steps of Section 1.5 of the book, i.e.:
    • Identification of input parameters,
    • Identification of performance metrics of interest,
    • Identification of relationships between parameters and variables,
    • Identification of rules imposed on the simulation model to be constructed
    • Collection of data (parameter values from given input data, distribution fitting)
  2. Second, an Arena model has to be constructed from the solution of part 1. Here, the lectures on sections 11-14 will come in handy, as they show commonly used constructs in Arena. In the report, attention should be given to steps 3 - 6 of Section 1.5 of the book, i.e.:
    • The model has to be designed according to the earlier specification,
    • The model should be validated. As empirical data is not available, sanity checks should be used.
    • Simulation experiments should be defined to solve the problems presented in the assignment.
  3. Third, the simulations should be run, the output collected and analysed and a final report should be presented providing the proposed solution to the given problem.