Loading...

About us

Decision Systems Lab

The Decision Systems Lab is a research entity in the Faculty of Engineering and Information Sciences.It is an interdisciplinary group of academics, post-doctoral researchers and research students drawn from the the School of Computing and Information Technology and the SMART Infrastructure facility at the University of Wollongong.










laptop



Research Themes

  • Medical informatics (under the aegis of the Centre for Oncology Informatics)
  • Data science
  • Business-process management
  • Service-oriented computing
  • Requirements engineering
  • Conceptual modeling
  • Agent-Oriented Systems
  • Constraint Programming
  • Knowledge Representation and Reasoning
  • Knowledge Management





DSL Centers

Need advise on a Technological problem

Get in touch with our experts

Our Directors

A/Prof. Hoa Dam

Associate Director

Associate Professor in the School of Computing and Information Technology, University of Wollongong (UOW) in Australia.

More details

Dr. Janusz Getta

Associate Director

Senior Lecturer, School of Computing and Information Technology, Faculty of Engineering and Information Sciences, University of Wollongong

More details

Prof. Aditya Ghose

Director

Professor of Computer Science in the School of Computing and Information Technology, University of Wollongong (UOW) in Australia.

More details

Case Studies

The Problems We Solved

IBM RESEARCH


The Problem

Service teams were being over-staffed – an easy way to deal with the prospect of paying penalties associated with SLA violations if leaner teams were used.


The Solution

Build a business case by using agent-based simulation of the service delivery setting. Build an optimal service despatcher using distributed constraint optimization (DCOP) technology. Use the agent-based simulation coupled with the optimized service despatcher to establish the savings that accrue.

IBM Research >>

XEROX RESEARCH (CONDUENT LABS)


The Problem

Leveraging past execution histories to predict cloud process performance under various cloud load scenarios.


The Solution

Build a tool that permits us to query past execution histories as if they constituted a causal theory correlating the process context (cloud load), process design and process instance-level data with process performance.

Conduent Labs >>

INFOSYS LABS


The Problem

Business processes are often described, modelled and executed at a level of granularity that precludes strategy-level analysis. Critical business processes would thus not figure in strategic planning and deliberations, despite having significant strategic impact.


The Solution

A novel notion of goal orchestration was developed. Using a client dataset with 65,000 data points (relating to IT incident management processes), it was shown that a variation of standard process mining techniques could be used to mine goal orchestrations.

Infosys Labs >>

IBM RESEARCH


The Problem

IBM and its clients routinely used the notion of “process variant” but had no concrete means of deciding whether a given process design or instance was a valid variant of another. This had significant implications for process compliance management.


The Solution

A bespoke framework (GOVM) was developed that used a machinery for extracting/identifying goals from process designs and instances to make a principled determination of whether a design or instance was a valid variant of another.

IBM Research >>

ADVANCED SYSTEMS INTEGRATION PTY LTD


The Problem

Truck despatch systems for open-pit mines help decide which truck should service which excavator (shovel, dragline etc.) but these tend to work in batch mode. They are thus not responsive to on-the-fly changes (roadway blockages, landslips, truck breakdowns etc.).


The Solution

An agent-based market-oriented solution where individual truck agents bid for work (the excavator agents are auctioneers) and subsequently internally trade jobs to maintain near-optimal performance.

CSC (BHP-IT)


The Problem

BHP Building Products’ manufacturing facility in Chullora used sub-optimal production scheduling strategies.


The Solution

A novel system called ISSUS was developed that used constraint programming technology at the back end for optimization, and offered, at the front end, an interactive Gantt chart that permitted operators to interrupt and alter system-generated production schedules while exploring the implications of relaxing applicable constraints (and generating what-if scenarios).

BLUESCOPE STEEL


The Problem

Long-term planning, medium-term planning and scheduling and short-term scheduling was happening in disparate silos, leading to locally optimal but globally suboptimal solutions. The conjoint problem was very large, involving upwards of 2.5 million decision variables.


The Solution

A novel problem decomposition scheme called hypertree decomposition was developed which permitted decomposed sub-problems to be solved separately while still permitting these solutions to be composed into a global solution.

Bluescope Steel >>

NSW STATE EMERGENCY SERVICES


The Problem

A vast operational infrastructure that remains dormant in normal times but “wakes up” during emergencies. A vast pool of volunteer staff. This infrastructure was not adequately modelled nor adequately understood.


The Solution

A very large-scale enterprise process architecture built using the i* notation. A novel form-based technique for model elicitation was also developed. These outcomes were put into routine used by SES managers and eventually informed the design of a system called RFAOnline.

NSW State Emergency Services >>

BLUESCOPE STEEL


The Problem

Real-time scheduling was being done in separate silos for the blast furnace, stockyard (for iron slabs) and the hot strip mill.


The Solution

A novel distributed constraint optimization technique called SBDO was developed and tested at Bluescope Steel.

Bluescope Steel >>


7
Awards

15
Trusted Clients

4,000,000
in Grants

Contact Us

Leave us a word

Our Office

Come over and have a cup of coffee!

Address

Decistion Systems Lab, UOW
6.208 SMART Infrastructure Facility
Northfield Avenue
Keiraville NSW 2522