Adaptive Supply Networks

Scientific Research

In addition to our industry partnership research we carry out strategic scientific research into topics relevant to supply networks.

Decision Making under Uncertainty

Much of our strategic scientific research effort is concentrated on the topic of decision making under uncertainty.

Statistics Research

Local Modelling is a term we use to refer to statistical methods for fitting models with large numbers of fixed effects (which are expected to remain constant, but need to be estimated), dynamic effects (which are also to be estimated but are expected to change slowly over time) and white noise (which is uncorrelated, like sampling and measurement errors).

The major project within which we are currently looking to apply our local modelling techniques is centred on Signal Analysis for Aluminium reduction cells: the analysis of voltage and current signals from the cells. Many quantities are measured on reduction cells, but many of these quantities are only measured at low frequency or with poor precision. It is crucial to extract as much information as possible from the continuous measurement of cell voltage or cell resistance.

Our local modelling methods are likely to be useful for signal analysis of data from large sets of reduction cells:

• The fixed effects used might include, say, differences between classes of cell designs, the relationship between cell resistance and bath temperature, and chemical and physical characteristics of the alumina being used.
• The dynamic effects used in models might include the current bath depth in the cells, the current bath temperature, and sizes of the gaps between the anodes and the molten aluminium.

We are undertaking this work as part of the Light Metals Flagship. An ASN team member, Lwin Thaung, recently presented a poster on related work at the Australian Statistical Conference.

PhD Students

ASN team members are currently co-supervising several PhD students. Amongst our current and recently-completed PhD students are:

• Kaveh Nezamirad (PhD Student at Swinburne Univ., co-supervised by Simon Dunstall) submitted his completed PhD thesis at the end of August 2007: Modelling Collaborative Planning and Scheduling Scenarios that include Human and Computer Decision Activities.
• Mauro Bampo (PhD Student at Monash University, supervised by Mark Wallace) started his studies in early 2008 and is being advised by Simon Dunstall and Gaurav Singh.
• Dhananjay “DJ” Thiruvady (PhD Student at Monash University, co-supervised by Bernd Meyer and Andreas Ernst) is working towards completing his thesis in 2009. During the 2007-2008 FY he submitted one refereed conference publication.
• Luke Mason (PhD Student at Deakin University, co-supervised by Vicky Mak and Andreas Ernst). During the 2007-2008 FY he submitted a paper that has been accepted for publication by INFORMS Journal of Computing. Expected to complete in late 2009 or early 2010.

Page last updated February 10, 2009 11:43 AM.
ASN sub-site maintained by Simon Dunstall.