Complex Software Systems & Services


Welcome
Login
I forgot my pass
print
Job and Student Offers

Job Offers


At the moment we do not have any open job positions.

PhD Student Offers


PhD Scholarships for Australian and International Candidates.


Opportunities exist for undergraduate, graduate and postgraduate research projects and scholarships in the research areas of the centre. Sample PhD topics are listed below but the prospective candidates with own research topics in the are areas related to the IAT research programs and projects are welcome.

Details of a range of the PhD scholarships offered by Swinburne and externally are available on the Swinburne Research webpage

Modelling and understanding user experience in agent-based smart exchanges

As the intelligent agent technologies are being used in the context of a smart exchange, the issues of user experience when interacting with and/or using the agents will become critical. These issues range from the user experience when interacting with the agents to the user experience when using the agents to the experience the consumer may have in a futuristic scenario where the use of agent technology would be the norm. For instance, how would the customer behave when they buy an insurance policy from an agent? Here, the customer may directly negotiate with a (software) sale agent, or the consumer may receive advices from a (software) broker agent on which product to buy. A human consumer may also have indirect interactions with software agents if they have to make purchase decisions in a market where other purchasers can be software agents.

The problem of modelling and understanding the user experience in various contexts of consumer-agent interaction will be of main focus in this project. The student will leverage existing works in the area of negotiation support systems and agent-based automated negotiation to investigate this research problem and carry out theoretical and empirical studies to obtain evidences about user experience and consumer behaviour in the difference interaction scenarios.

Skills Required:
One or more of the following (and willingness to learn): Essential – problem solving and analytical skills, algorithm design and Java programming; Desired - understanding of (one or more) simulation (e.g. Matlab), artificial intelligence, negotiation and decision optimisation techniques, intelligent agents and multi-agent systems, mechanism design, behavioural economics.

Intelligent data analytics-driven monitoring and management of cloud service-based applications

Modern cloud service-based applications (CSBAs) (e.g. multi-tier enterprise applications, social networking applications, environmental modelling and other data-intensive applications) are increasingly being composed from multiple components that require and consume cloud services at different layers of the application stack. A key feature of cloud computing that CSBA providers can take advantage of is elasticity i.e. the ability to respond to dynamic workloads and resource demands. Autoscaling allows the rapid provision and release of cloud services in an on-demand and automated fashion, giving cloud consumers the perception of unlimited cloud resources. In principle, CSBA providers can handle both dynamic workloads at the application level, and resource constraints and performance degradations at the cloud service level, with minimal management effort and provider interaction. However, in practice, this is not easy to achieve. CSBA providers need to know when to scale, where to scale and by how much to ensure quality-assured application provision. They also need to take into consideration the time taken to bring new capacity online.

This project addresses a key challenge of the quality-assured provisioning of CSBAs. It aims at developing an innovative framework that takes an analytics-driven approach to monitoring and management of CSBAs. It will develop new models, mechanisms and tools for the agent-based quality-assurance of CSBAs.

Skills Required:
One or more of the following (and willingness to learn): Essential – problem solving and analytical skills, algorithm design and Java programming; Desired - understanding of (one or more) cloud computing, data analytics, artificial intelligence, multi-agent systems, quality-of-service.

Adaptive Process Automation in Cloud Manufacturing

Information technologies (including cyber-physical systems, the Internet of Things – IoT, and cloud computing) have been identified as the main enablers for future manufacturing generations. To ensure that the enabling technologies work well together and for information to be effectively acquired and shared among the decentralized collaborators, a management layer is required. This management layer integrates various business models across different collaborators, facilitates resource and capability discovery and composition and support critical functionalities such as quality assurance, information security and data compatibility. In this project we aim to develop a prototype for this management layer with a specific focus on the quality assurance functionality with the support of various industry standards and the development of a management framework for decentralized collaborators based on Service Level Agreements (SLAs).

This project aims to develop new agent-based techniques and software tools for the adaptive service agreement and process management in order to ensure coordination and adaptive provision of complex manufacturing processes with end-to-end Quality of Service.

Skills Required:
One or more of the following (and willingness to learn): Essential – problem solving and analytical skills, algorithm design and Java programming; Desired - understanding of (one or more) smart grid, artificial intelligence, multi-agent systems, peer-to-peer systems, game theory.

Socially Intelligent Demand Control in Smart Grids

This research project will investigate the demand control of electric power through the utilisation of socially intelligent computing, i.e. peer-to-peer interactions among consumers (represented and supported by intelligent software agents) to guide their individual control decisions and, by aggregating the decisions, produce effective control of system demand that minimises their collective greenhouse gas footprint. In general terms, the problem involves the allocation of scarce societal resources (i.e. electric power) among independent consumers that have private and often conflicting preferences, which cannot be controlled centrally in order to satisfy the global objectives (i.e. low carbon footprint). Instead the projects will focus on decentralised control that involves two levels of peer-to-peer interactions and optimisation among the participants. At the macro-level, interactions are needed to achieve consensus on the overall goals and trade-offs among alternatives, which would produce analogues of supply-and-demand curves. At the micro-level, interactions are needed to cause individual decisions to be as "far apart" or dissimilar as possible so as to spread demand as uniformly as possible. The objective of this research is to study and develop the efficient consensus and incentive-based mechanisms for decentralised demand control that respects system-wide social objectives and individual preferences. The mechanisms will influence consumer decisions regarding local energy usage, generation, and storage, as well as overall energy supply and demand, according to local consumption preferences of individuals and system demand objectives of grid operators.

Skills Required:
One or more of the following (and willingness to learn): Essential – strong analytical skills, algorithm design and Java programming; Desired - knowledge of (one or more) smart grid, simulation, optimisation techniques, peer-to-peer systems, intelligent agents, market-based mechanisms, game theory.

Collective Optimisation of Resource Allocation in Smart Infrastructure / Smart Grids

Smart infrastructure encompasses networked infrastructure that uses ubiquitous sensor, information and communication technologies to better utilise or sustain resources. Examples of smart infrastructure include electricity grids that improve grid reliability and better utilise energy, transport systems that optimise traffic flows, and water networks that improve water productivity in agriculture. This research will investigate the mechanisms of collective optimisation of resource allocation in smart infrastructure considered as a socio-technical distributed dynamical system. Distributed dynamical systems are formed by a number of autonomous entities (human and software agents) that control their own decisions and behaviour, usually with little direct communication and interactions. The overall behaviour and performance of the system depends on the individual agents' behaviours and interactions with the environment that often results in an emergent behaviour of the system. The objective of this research is to study and devise efficient mechanisms and decision strategies based on the principles of inductive reasoning and bounded rationality for system-level optimisation of shared resource utilisation resulting from the collective decisions of independent agents without direct communication between them in socio-technical distributed dynamical systems, such as smart infrastructure in general and smart energy grids in particular. The mechanisms will be applied and evaluated in smart energy management to optimise the distributed electric vehicle charging so the individual needs, preferences and constraints of the participants are satisfied (e.g. time and cost preferences, battery characteristics etc), while the goal and constraints of the system are met (e.g. overall load limits, social fairness requirements, etc).

Skills Required:
One or more of the following (and willingness to learn): Essential – strong analytical skills, algorithm design and Java programming; Desired - knowledge of (one or more) smart grid, simulation, optimisation techniques, multi-agent systems, inductive reasoning, dynamical systems, game theory.

Self-learning Market-based Resource Allocation in Smart Infrastructure / Smart Grids

Smart infrastructure encompasses networked infrastructure that uses ubiquitous sensor, information and communication technologies to better utilise or sustain resources. Examples of smart infrastructure include electricity grids that improve grid reliability and better utilise energy, transport systems that optimise traffic flows, and water networks that improve water productivity in agriculture. The emerging smart infrastructure presents a major paradigm shift in resource allocation management with the aim to extend the centralised supply management towards the decentralised supply-and-demand management that will enable more efficient, reliable and environment-friendly utilisation of resources. The objective of this research is to study and develop a theoretical and practical framework and mechanisms for self-learning market-based resource allocation with strategic reinforcement learning agents to simultaneously optimize both the individual and social allocation efficiency of distributed resources in smart infrastructure. In particular the developed strategic supply-and-demand mechanisms will be applied and evaluated in market-based energy management in micro-grids (involving local consumers, prosumers and producers) to optimise the energy used within the neighbourhoods, mutli-occupancy buildings and/or precincts.

Skills Required:
One or more of the following (and willingness to learn): Essential – strong analytical skills, algorithm design and Java programming; Desired - knowledge of (one or more) smart grid, simulation (e.g. GridLab-D), optimisation techniques, multi-agent systems, reinforcement learning, market-based mechanisms, game theory.

Human and Software Agents Negotiation

Both human and software agents engage in interactions with purpose to acquire products or services, undertake joint activities, achieve common goals, and share knowledge and information. Due to different and often conflicting preferences, objectives and information those interactions need to include negotiations and conflict resolution. This project will develop, evaluate and determine efficient and effective mechanisms for negotiations and conflict resolution involving human and software agents in various exchange configurations, interaction types, role assignments and information environments. The ultimate goal is to develop an adaptive e-negotiation software tools that can provide most efficient and effective mechanisms according to individual needs and abilities of the human and software agents.

Skills Required:
One or more of the following (and willingness to learn): Essential – strong analytical skills, algorithm design and Java programming; Desired - knowledge of (one or more) simulation (e.g. Matlab), artificial intelligence, negotiation and decision optimisation techniques, intelligent agents and multi-agent systems, mechanism design, behavioural economics.

Strategic Learning Agents in Equilibrium-based Markets

Equilibrium-based market mechanisms in multi-agent systems offer efficient solutions for distributed coordination and resource allocation required in a range of application domains including computer networks, transportation and energy management. The objective of this research is to devise strategic learning agents and investigate their impact on the individual and social outcomes of market mechanisms based on general equilibrium theory. It involves the application of AI techniques, such as machine learning, to develop the agents’ strategic behaviour, and the theoretical and empirical evaluation of the outcomes of such strategic participants in the market-based systems.

Skills Required:
One or more of the following (and willingness to learn): Essential – strong analytical skills, algorithm design and Java programming; Desired - knowledge of (one or more) market-based mechanisms, multi-agent systems, machine learning, game theory.

Coalitions for Responsive Demand Management in Micro-grids

The project will develop decentralised, agent-based mechanisms for dynamic coalition formation in micro-grids so that agents representing each home can self-organise into relatively stable groups that coordinate their electricity use. Dynamic coalition formation is decentralised, responsive and not dependent on prediction, making it a natural choice for this problem, as the problem space includes unpredictability and requires near real-time responsiveness to changing conditions. This approach will facilitate highly responsive demand management, offsetting the need for spinning reserve and lowering the cost of electricity production while reducing its environmental impact. Micro-grids with an ability to internally manage their electricity needs could more efficiently manage renewable electricity generation while helping to integrate the charging of electric cars into our energy infrastructure, protecting grid stability. They could operate with less dependence on the rest of the electricity grid, reducing the need for external generation and transmission.

Skills Required:
One or more of the following (and willingness to learn): Essential – strong analytical skills, algorithm design and Java programming; Desired - knowledge of (one or more) smart grid, artificial intelligence, multi-agent systems, peer-to-peer systems, game theory.


For further information and application forms, please contact Prof. Ryszard Kowalczyk with the following details:
  • A detailed curriculum vitae
  • An (electronic) copy of undergraduate and postgraduate transcripts
  • Topic(s) of interest preferably with a very short synopsis (max half a page)
  • Evidence of formal research experience if applicable (e.g. degree by research, research projects, publications)
  • Evidence of English proficiency (native English speaker, previous studies in English in certain cases, or IELTS of 6.5 with no band below 6)

For more info also see:

Visits by Postgraduate Students and Postdoctoral Fellows

Opportunities for international postgraduate students and postdoctoral fellows exist for extended research visits in the centre.

For further information , please contact Prof Ryszard Kowalczyk (see below for contact) with the following details:
  • A detailed curriculum vitae
  • An (electronic) copy of undergraduate and postgraduate transcripts as appropriate
  • List of publications and research achievements
  • Topic(s) of research interest with a very short synopsis (max half a page)

For more details (funding, application forms etc) see:

© Swinburne  Copyright and disclaimer information
last modification: Monday 17 of October, 2016 [00:27:03 UTC] by rkowalczyk
Maintained by: IAT Webmaster (mchhetri@swin.edu.au)