Defence Researches and Innovations
1) Predictive Big Data Analytics for Logistics Operations Management (Prof E. Chang, Dr O. Hussain, Dr N. Janjua and A/Prof G. Heaslip )
Logistics systems in today’s world are characterized by increasing dynamicity which arises from the trends of global economy. Defence & Business are no exception to this trend & require high quality decision-making support systems to optimize their logistics operations & meet the challenges. Big Data provides a wealth of information from which data can be analysed real time and knowledge synthesized from it. However, before this can be achieved appropriate tools & techniques are needed first to understand the semantics of Big Data, integrate them & apply predictive analytics on it. In this project, we develop the foundation blocks to achieve this aim.
2) Defence Logistics Information Systems - Ensuring interoperability across Army, Navy and Airforce (Prof E. Chang, Dr. A. Talevski, Dr N Janjua, Dr D Prior (SBUS), A/Prof R Sarkar (SEIT) and Prof T Dillon )
Defence logistics underpin Defence capability. An essential component of successful logistics systems is interoperability – the capacity for diverse systems to work together. However, diverse and separately developed information systems currently comprise the Defence logistics approach. Army operates the Military Integrated Logistics Information System (MILIS). However, the stove-pipe nature of MILIS resulted in interoperability issues with other systems in defence such as SLIM (Ships Logistics Inventory Management System) and Air Log (Air force logistics systems). All three systems are independently auditable. However, there are no synchronised support functions to allow unified inventory and asset tracking. This project seeks to address this issue.
3) Creating Intelligent Situational Awareness: Towards Human-Centred Data Mining and a Recommender System in Army Logistical Environment (Prof E Chang, Dr D Prior, A/Prof G Heaslip and Dr F Hussain (UTS))
Data mining as a core capability is an intrinsic part of every intelligent organization and enterprise since 9-11 terrorist attacks in the USA in 2001. It is used not only for counter terrorism cyber security, fraud detection, but also as a core technology is used in domains, including commerce and science, such as business intelligence; human disease studies, energy, global warming and social networks. One of the core objectives of data mining is to discover new and useful patterns from data that gives rise to discovery of new and previously unknown information and knowledge. In this project we develop Data Mining techniques to achieve the Intelligence in Defence applications.
4) Risk-based decision support for Asset Management in land force sustainment (Dr O Hussain, Dr N Janjua (SBUS), Prof P Hyland (QUT) and Prof E Chang )
Defence assets comprise approximately $40.3 billion of specialist military equipment, $21.1 billion of land building and infrastructure and $5.7 billion of inventory. These significant financial expenditures along with some of the recent incidents that demonstrate poor Risk management in Asset sustainment practices brings with it questions regarding Australian Army’s effectiveness in establishing, implementing, and improving risk management practice in its assets sustainment in land force operational support. Best practices on Asset Management and ISO standards in recent years have provided with a Total Life Cycle Asset Management (TLAM), which stresses on a clear definition of risk measurement dimensions and risk prediction techniques in the context of Asset management. The objective of this research project is to develop such a cognitive tool that will eventually lead the Army in better risk management of their assets in the future.
5) User-side Quality of Service (QoS) management of Cloud-based Services (Dr O Hussain)
The aim of this project is to develop a Risk Assessment as a Service (RaaS) module that enables the service users of Cloud Computing service to make informed service-based decisions with the service providers. Developing such a framework will address one of the major obstacles of Cloud Computing mentioned in the literature. A prototype system of the proposed framework will be developed in this project that can be applied and tested in situ
6) Big Data Analytics through Conjoint Data and Content Mining(Prof E Chang, Prof TS Dillon, Dr A. Talevski, Dr N Janjua and Dr O Hussain )
Large amount of data is collected and stored by various, government, industrial, commercial or scientific organisations and consists of (1) structured data, eg Asset Databases and (2) semi-structured/unstructured content such as customer complaints. As the complexity and volume of the data continue to increase, the task of classifying new unseen data and extracting useful knowledge from the data is becoming practically impossible for humans to do. This project develops a methodology for performing advanced analytics from both data and content conjointly. The techniques developed will be used to address important issues in Defence enterprise, business organizations and health medicine domain.
7) Green Logistics based on Cyber Physical Systems for Minimizing Emissions in Road Transport and Logistics(Prof E Chang, and T. Jensen )
The transportation system contributes 18% of the nation's carbon emissions. Diesel powered logistics vehicle are a major contributor to this. Conversion of these diesel trucks to LNG (Liquefied Natural Gas) will help achieve better economies and emissions profiles. This poses challenges and requires strong evidence to convince truck companies of the benefits. This is achieved by real time monitoring of the Diesel to LNG converted trucks in actual operation and different contexts. This information is then aggregated in a repository using a web enabled Cyber Physical System. This is analysed using knowledge discovery techniques. This knowledge is used to develop and design an intelligent controller for the LNG/Diesel converted trucks.
8) Intelligent Asset Management( Prof E Chang, Dr O Hussain, and Dr N Janjua)
This project investigates and develops techniques for Intelligent Asset Management. It mines and tracks multiple information resources for accurate Asset information sharing, knowledge synthesis and financial accountability through the use of Ontologies and XML profiles. Novel techniques for conjoint mining of structured and semi-structured data and management. The results will be validated using large real-world data sets provided by the partners.
9) Next Generation Data Merging, Cleansing, Prevention of Data Pollution for Big Data and Complex Data using information from multiple sources(Prof E Chang, Prof TS Dillon, Dr. A. Talveski)
UNSW and MIMOS has been successfully established research collaboration in the area of merging and cleansing of large real databases for PERKASO/SOCSO (Social Security department of Malaysia). There are a number of research issues which must be addressed for the development of the next generation of Data Merging and Data Cleansing Techniques and Tools. This project is designed to address these issues to allow MIMOS to be at the cutting edge of this highly significant Technology and move MIMOS into a world leading position in Data Merging and Data Cleansing.
10) Cyber Physical Systems for Intelligent Asset Tracking in Defence Logistics(Prof E. Chang, Dr. A. Talveski)
In this project, the aim is to develop such an advanced Cyber Physical Systems for Intelligent Asset Tracking in Defence Logistics. Specifically it will develop a Cyber Physical Systems with enhanced Internet of Things, Auto ID Systems, Digital Pens and Smartphone Technologies for Intelligent Asset Tracking and Provenance of Goods and Assets for the future Australian Defence Force. It will result in considerable efficiency, productivity, capability development and lead to benefits for the Australian Defence and the commercial logistics sector as well.
11) A Framework for Risk Identification and Mitigation in Transport Logistics for Homeland Security(Prof E. Chang, Dr. Omar Hussain, Dr N. Hyndman-Rizk)
In this research, we intend to develop such a framework on the identifying, determining, assessing, mitigating and monitoring the source of risks, with the following 5 major steps:
- identifying the risk sources in the logistics network;
- determining the probability of failure or probability of the occurrences of those risks that may lead to failure in the logistics network;
- ascertaining the adverse consequences in the logistics network as an occurrence of those risks;
- assessing the impact of losses on financial, human and resources;
12) A Co-joint Data Mining Approach for Minimising Emissions(Prof E. Chang, Prof TS Dillon, Prof S. Chand, Dr. A. Talevski,)
This research will address the issue by defining:
a) type of subtrees to be mined (embedded, induced or ordered)
b) support definitions or constraints (transaction based support, occurrence-match support or hybrid-support) depending upon the aims of the query and the way how the information is organized in the augmented database.
c) The association rules that are to be synthesised, extracted knowledge pattern should be analysed for their interestingness and relevance to remove any redundant, misleading and irrelevant rules.The aim is to increase the confidence in the determined association rules being a true reflection of the patterns that exist in the underlying augmented data.