University of Glasgow
Investigating the direct application of chaos theory to detect, analyse and anticipate high-level variability in the logistics demand of third party logistics
Abstract
dc:description.abstractThird party logistics providers operate in an environment where the customer demand shows high-levels of variability. Existing methods of analysis and prediction cannot capture these types of fluctuation. Newer methods of analysis, and the identification of potential chaotic conditions to explain these intense oscillations, have not been tested for applicability in these situations. The purpose of this thesis is to investigate whether the direct application of chaos theory can efficiently detect, analyse and anticipate high-level variability in the logistics demand of third party logistics (TPL). The research involves a single case study analysis. The variable investigated is the logistics extracted from the EDI files of the company in a time-series format. The time scale of the data is over two years. A framework of data analysis, called CASTS (Chaotic Analysis of Short Time Series), is constructed in order to analyse the data. It is an amalgamation of linear, non-linear and chaos theory based techniques selected to allow the detection, analysis and possible anticipation of the underlying data set. The CASTS method is composed of the application of the autocorrelation function, power spectrum, BDS statistics, mutual information, phase space plots, correlation dimension, Lyapunov exponent and finally, Hurst exponent tests. In addition a surrogate data test is performed in order to achieve a 95% level of confidence in the results. The importance of this research is fourfold. First, it proposes a solution for third party logistics to improve their operational efficiency through an enhancement of their forecasting, planning and control abilities. Secondly, it adds new knowledge to logistics management in two ways; it brings together two different sciences and provides insights that have not been explored before and; it succeeds to identify, for the first time, the presence of chaotic behaviour in real logistics data and thus give a new direction to logistics research. Thirdly, it provides CASTS as a new framework of analysis for the detection of chaotic behaviour in short time series that was not previously applied in social sciences. Finally, it has tremendous implications for industry; it concerns the logistics anticipation, planning and control. It assists companies to focus their efforts in understanding the structure and restraining the behaviour of their demand patterns rather than focusing in reactive actions.
Degree
thesis:*- Level dc:type.qualificationlevel
- PhD
- Grantor dc:publisher.institution
- University of Glasgow
- Year dc:date.issued
- 2001
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Papadopoulou, Chrisoula
Subjects
dc:subject × 2Rights
- Language dc:language
- en