Illustration: Ken Uchida

Ten years ago, the internet was still new technology and it was enough that it opened up unprecedented access to vast amounts of information at the click of a mouse.

Fast-forward a decade and in the increasingly ubiquitous world wide web people are demanding greater sophistication in the way they interact with the ‘net’ – be it to book a holiday, manage personal finances, or monitor a complex supply-chain system in their business.

To meet this demand, a new paradigm of computing – service-oriented computing – has emerged to better integrate web services and, in so doing, make them capable of handling increasingly complex requests that might rely on real-time information from several sources at once.

Professor Ryszard Kowalczyk, director of Swinburne University of Technology’s Centre for Information Technology Research (CITR), draws on an example from within the travel industry to explain how the new model could work.

"Instead of going to a travel agent or its website to find and book a flight, then adding accommodation and a car separately, you can simply book a personalised holiday package as a dynamically integrated service which has different services provided by a range of different providers, all from a single request."

Professor Kowalczyk expects such service-oriented computing platforms to be employed in a broad range of industries and organisations, especially the management of supply chains and logistics where a number of different entities are typically involved.

Another researcher at the CITR, Professor Jun Han, says that while the integration of various software and enterprise systems can be done manually, service-oriented computing provides a model in which this integration process can be automated.

"It involves a set of standards, and is really, if you like, a new way for systems to interact with each other," Professor Han says.

With a large portfolio of research projects and more than 30 researchers developing new service-oriented technologies, CITR at Swinburne has become one of the largest concentrations of research expertise in service-oriented computing in Australia.

It is involved in a range of substantial projects including five funded by the Australian Research Council (ARC), two funded by the Department of Education, Science and Training and several collaborative projects with industry partners. CITR is also a key partner in the recently approved national Cooperative Research Centre for Smart Services.

The projects include one led by Professors Kowalczyk and Han that recently resulted in the creation of the Adaptive Services Agreement and Process Management (ASAPM) platform. This can not only connect a range of different web services on demand, and provide a single user interface through which they can be accessed, but can also adapt to changes and handle faults that may occur while using web services, ensuring the requested functionality and quality of services.

Three industry partners – telecommunications provider Telstra, the Australian Government’s Defence Science and Technology Organisation and internet multimedia company Everyday Interactive Networks – were involved in the project to ensure the platform was capable of handling the requirements of a broad range of organisations and industry sectors.

The ASAPM project also formed part of the European Union’s Adaptive Services Grid, a larger project involving 21 companies and universities including Swinburne – the only Australian partner – which also aimed to develop service-oriented architecture for complex service delivery.

Another project in which CITR is involved includes a joint industry and ARC-funded initiative with CA, one of the world’s largest independent software companies.

With a focus on the integration of the internal systems of large, multinational companies so that data and capabilities can be managed effectively and efficiently, the project investigates the next generation of web service registries – a technology that performs a function similar to a Yellow Pages® for software services within a large company or organisation.

While such directories are presently no more than just that – a listing of where particular services can be found – the project is looking at how these directories can provide access and management to a range of these different services at the same time.

Tim Ebringer, a Melbourne-based researcher at CA Labs, an R&D business unit within CA, says the project, which has been running for a year and will run for a further three years, is about making the registry "active instead of passive".

"We go out to the enterprise and check that the services are working according to expectation and manage the lifecycle of web services through development, beta testing, pre-production and production," he says. "We want to understand and create a web service registry that becomes the focus of an enterprise’s service-oriented architecture strategy."

Professor Han, who leads the project at CITR, says that such ‘rich’ service registries "will be the centrepiece of future enterprise information systems deploying the service-oriented architecture.

"They are ‘rich’ both in handling the services’ functionality and qualities, and in managing services and service-based applications throughout their lifecycle," he says.

Professor Kowalczyk sees the project as another case of how service-oriented computing platforms can deliver targeted information and functionality to particular audiences in increasingly sophisticated ways.

"It can become quite a complicated process – it’s basically what we do in real life and beyond," he says, returning to the travel example. "We negotiate with a number of (travel) providers at once and then choose a combination, so that together they satisfy our requirements in the best way."