A Generator for Volatile Demand Profiles. A Brief Description of a Tool


Essay, 2017
9 Pages, Grade: 0,0

Excerpt

1 Introduction

To test, compare and improve algorithms and new approaches for staff scheduling, planning problems are needed. Such problems are often not available in literature. Therefore, the Demand Generator was developed to create own problems for staff scheduling. The actual parameters of this tool are based on real data of a call center, which were only slightly modified for the purpose of data protection of the call center. Nevertheless, the tool can also be used for other industries like retail sector or logistics.

The tool creates for a whole year in hourly resolution values for the staffing demand of a specific workstation or a specific function. Opening hours, holidays, weekdays, min/max staffing levels, stochastic elements, events and weather effects are taken into account by the Demand Generator. Fig. 1 shows an example of the calculated demand for one week.

illustration not visible in this excerpt

Fig. 1: Final demand of the Demand Generator for a week in hourly resolution.

The Demand Generator is public available and it is allowed to modify it. If the tool or a modified version of the tool is used for a publication, reference should be made to this working paper:

Günther, M.: A Generator for Volatile Demand Profiles, Working Paper 07/17, Munich: GRIN, 2017.

The Demand Generator was developed with MS EXCEL™ and Visual Basic. It can be downloaded at:

http://www.maik-günther.de/DemandGenerator.xlsm
If there are any problems with the ‘ü’ in the URL, please use the following link:
http://www.xn--maik-gnther-yhb.de/DemandGenerator.xlsm

1 Function of the Demand Generator

The Demand Generator was developed in MS EXCEL and Visual Basic and works in the following way:

1. A deterministic core of the hourly demand is given. These values can be modified by the user of the tool. Without any other modifications this demand is used for each day of the year. Opening hours are taken into account if a 0 is used for the demand. Fig. 2 shows an example for the deterministic core with opening hours from 8 AM till 6 PM.

illustration not visible in this excerpt

Fig. 2: Deterministic core for 24 hours of each day.

2. There are often differences between the overall demand of weekdays. If a shop or a call center is closed on Sunday weekday factors can be used to set all demand values to 0 at each Saturday of the year (here a weekday factor of -100). Other effects like an increased number of calls on Mondays in comparison to Fridays can also be modeled with weekday factors. Based on the weekday factors the deterministic core will be changed for all days of the year. Fig. 3 shows an example of weekday factors. These values can be modified by the user.

illustration not visible in this excerpt

Fig. 3: Weekday factors.

3. The Demand Generator can also use events which influence the demand of specific days. Such events are often planned by the marketing departments. Other types of events are for example holidays, weather effects, important football games, etc. The user is free to create own events with own specific effects. Examples are given in Fig. 4.

illustration not visible in this excerpt

Fig. 4: Definition of events.

After events are defined, they can be assigned to the days of the year. It is possible to assign more than one event to a day. Fig. 5 shows an example. Here the event P1 (the first day of a price adjustment) is assigned to the day with the number 337. As a result the demand from step 2 will be modified according the assigned events.

A Generator for Volatile Demand Profiles

illustration not visible in this excerpt

Fig. 5: Assignment of events to days.

4. To take stochastic effects into account an Ornstein-Uhlenbeck process can be added to the result of the previous step. The user can choose own parameters for the Ornstein-Uhlenbeck process. The preselected parameters are presented in Fig. 6.

illustration not visible in this excerpt

Fig. 6: Parameters of the Ornstein-Uhlenbeck process.

5. If the hourly values of the demand should be limited, an upper and lower bound can be given (see

Fig. 7 for an example).

illustration not visible in this excerpt

Fig. 7: Min/max levels of the final demand.

Checkboxes are available to activate or deactivate the steps 2-5 (see Fig. 8). It is possible to deactivate all checkboxes and to run the model. The result will be the deterministic core at all days of the year without any modifications.

illustration not visible in this excerpt

Fig. 8: Check boxes and the start button of the demand generator.

The output of the calculation is given in two ways.

- One table with 168 columns (hours from Monday till Sunday) and 52 rows (weeks of a year).

[...]

Excerpt out of 9 pages

Details

Title
A Generator for Volatile Demand Profiles. A Brief Description of a Tool
College
Technical University of Ilmenau
Grade
0,0
Author
Year
2017
Pages
9
Catalog Number
V373555
ISBN (eBook)
9783668513181
ISBN (Book)
9783668513198
File size
912 KB
Language
English
Tags
staffing demand, workforce management, call center
Quote paper
Dr. Maik Günther (Author), 2017, A Generator for Volatile Demand Profiles. A Brief Description of a Tool, Munich, GRIN Verlag, https://www.grin.com/document/373555

Comments

  • No comments yet.
Read the ebook
Title: A Generator for Volatile Demand Profiles. A Brief Description of a Tool


Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free