The Distributed Systems Technology Centre is hard at work shaping the language that may finally unite Australia’s divided e-government initiatives.
Andy Bond has quite a job ahead of him.
As chief scientist at the Distributed Systems Technology Centre (DSTC), he’s playing a leading role in an ongoing effort to develop consistent interoperability standards for linking departments and organizations at all levels of government.
After nearly five years of research, technical development and consultation with various government stakeholders, the first tangible result of that work — the publication of National Office for the Information Economy’s (NOIE’s) Interoperability Technical Framework (ITF) for the Australian federal government — was delivered in June 2003. (NOIE’s responsibilities have now been split between the Office for the Information Economy [OIE] and the Australian Government Information Management Office [AGIMO] — more on that later.)
This year, the DSTC team is getting the chance to do it all again. Revisiting the initial document to incorporate strategic, policy and technical changes since the time it was first created, Bond’s team is helping to shape the vocabulary of the language that, it is hoped, will finally engender long-elusive unity between Australia’s highly factional e-government initiatives.
A key step in doing this, Bond believes, is throwing out much of the conventional wisdom about enterprise application integration (EAI). Born of preconceptions that the proper role of EAI was to connect everything to everything, government and private sector organizations alike have pursued integration relentlessly, often to considerable expense, time and bother. Yet despite their work, introducing a new application into a traditional EAI environment — or even upgrading an existing one — means hard-coded application interfaces need to be retested and often rewritten.
“Rigid behaviour has been the compromise of integration,” Bond says. “EAI works very well in small systems, but the way to put a spanner in the works is to say ‘how is this going to work in a large system?’ A lot of integration solutions were built around distributed systems technologies, but don’t tend to be able to think of [resources] in a dynamic way. Predictability gives you confidence, but you lose the ability to react in a timely manner to leverage new business opportunities.”
That is no way to build the kind of expandable e-government infrastructure that today’s operating climate demands. Even as government bodies rework their citizen interactions to improve their ability to find and generate those opportunities, technical standards for standardizing methods of interaction have lagged far behind. The result: coders like those at DSTC work furiously to translate the broad edicts of their political taskmasters — for example, exhortations by Defence leaders that tomorrow’s information-driven soldiers should be able to access any type of data anywhere — into workable and broadly accessible systems.
Another oft-discussed beneficiary of better communications between applications is health-care, which has been struggling for years to facilitate the flow of information between the largely isolated fiefdoms of regional, state and Commonwealth health authorities. There, transfer of patient information between related bodies is more often than not conducted by health-care workers with a phone, fax and a niggling sense that there must be a better way to share such commonly needed information.
That “better way” has remained elusive, with the closest thing so far being the Commonwealth’s series of front-end Web portals, which guide visitors through the red tape necessary to handle a variety of life events. Yet while they provide functional integration — combining links to processes at different departments that each share responsibility for the particular task — portals have made no progress towards the broader goal of back-end interoperability.
Visions about a future where data flows smoothly may provide valuable guidance from a policy sense, but they are a nightmare when it comes down to actually delivering. Not the least of the problems facing government technical workers is the need to assure compliance with a quagmire of government policies and regulations — the double-handling and measured efficiency that ensure proper governmental operation but seem horribly redundant compared with the results-driven ethos of the corporate world.
Individual business leaders in a company, after all, may have reservations about helping the IT department implement an integration framework that shares their data. But when the CEO puts his or her foot down, those same business leaders will inevitably fall into line, even if only out of fear for their jobs.
In government organisations, however, the quagmire of potential political and technical pitfalls is much more tedious. Procedural isolationism is a way of life within many departments that have become so accustomed to operating independently that data sharing is sometimes inconceivable. Furthermore, such departments often lack the common sense of purpose that drives private business leaders to work together. Given that data exchange is a fundamental component of ubiquitous e-government, such attitudes represent a significant obstacle.
So how does DSTC, with a mandate to lay out common technical standards so that data isolationism becomes a thing of the past, even begin driving change through government?
First, by throwing out the term “integration” and replacing it with one it has been carelessly used interchangeably with in the past: “interoperability”. Integration, in this argument, describes a technical practice in which bridges between two systems overcome their incompatibilities. Interoperability, on the other hand, takes a step back from technical prescription and instead focuses on the real task at hand: communication.
The difference may seem subtle, but it is an important one for a project team that is charged with the task of sowing the seeds for carrots that will convince government departments to fall into line. After all, if the departments can exchange data in a meaningful way, who really cares what they are doing inside their four walls? Give them their sense of independence, as long as they can externally interface with interoperability frameworks like ITF.
Talk Shop, Not Tech
Building a culture of interoperability, according to DSTC senior research scientist Andrew Wood, has been difficult in an environment where discussion about inter-departmental communication has often focused more on risk management, allocation of liability and even pedantic issues such as the format of bits going down the wire.
The dangers of a technically prescriptive approach to interoperability become clear when considering governments’ past missteps towards standardization. A 1995 Multimedia Victoria discussion paper about GOSIP (Government Open Systems Interconnection Profile) policy, for example, talked seriously about the need for integration between departments and encouraged all state agencies to migrate from SNA or TCP/IP-based networks to a framework based on OSI standards.
Not long afterwards, TCP/IP and the Internet became the lingua franca of open systems and XML (eXtensible Markup Language) took the helm as the driving force for data description and interchange. The future of interoperability, it became clear, would rest on departments’ ability to standardize their information storage and publishing capabilities around these technologies.
Equally important is the ability to put technical details out of mind while negotiating the more important business processes and data flows that will be involved; technology will follow from there. “You can’t say something like ‘thou shalt use XML’,” Bond says. “That doesn’t really take you anywhere.
“Ideally, what we want to see come through in the framework is a set of best practice descriptions about how organizations can use standards. We often get driven not by the standards but by the technical implementations of those standards — but we want to be able to sell new standards and move organizations on to new ways of doing things.” Stepping away from technology specifics is one of the goals of the review of ITF; the second version, Wood hopes, will be more effective at guiding government departments through the broader issues around interoperability.
Where the first version of ITF provided the vocabulary for interoperability, the revised version will provide the grammar and context. Wood envisions the inclusion of several best practice case studies highlighting successful interoperability efforts that have already been executed by banner departments such as the Australian Taxation Office (ATO) and the Health Insurance Commission (HIC). Other mooted features include a pool of experts available to provide advice on interoperability, and best practice guides for addressing common government information management issues.
“There’s a soup of technologies and standards available,” Wood says. “The [ITF] is a good catalogue of standards, techniques and rationales — but the feeling is that it needs to be made a bit richer in terms of how to use the standards and technologies without having to think about the technical details. Departments understand the importance of interoperation, but there needs to be a deeper understanding of its capabilities.”
CIOs Have Their Say
Given their inevitable importance in the implementation of interoperability frameworks, CIOs will need to take on an advocacy role with respect to the ITF and its application to the idiosyncrasies of individual departments. In recognition of this fact, the NOIE-backed CIO Council — reporting to the Information Management Strategy Centre (IMSC), which has broad oversight of government information management processes — has taken a proactive role in applying the spirit of the ITF to real-world business problems.
Through a series of working groups, CIO Council members are nutting out the issues and obstacles raised by interoperability requirements across a variety of government information management requirements. Focus groups include the Commonwealth Employee Identity Management Group (CEIMG), Channel Management Working Group (CMWG), Authentication Working Group (AWG), and Attorney-General’s Online Verification of Identity Documentation (OVID) Working Group.
Each of these groups is seeking to apply the philosophy of e-government integration to the very specific business issues around identity management, identification of and communication with suppliers, confirmation of employee identification, and other related issues. By setting an agenda for addressing these issues within a government context, these groups should ultimately be able to establish best practice guidelines that will help other departments meet common benchmarks for functionality and effectiveness. This must all happen independently of DSTC’s technical work.
Framed in contemporary technology terms, it is tempting to say that these efforts could result in a compendium of Web services providing secure authentication and identity management services across departments. Waxing lyrical about the importance of ubiquitous Web services may sound like a respectable goal on the surface, but Wood is quick to remind that it is important for government to walk before it tries to run.
“The immediate goal is not just an open sea of services that agencies are able to dynamically pull together to achieve some functionality,” he says. “That’s a nice goal, but it’s way too ambitious for organizations’ structures to achieve at the moment. The realistic goal on this is point-to-point [communication], or by the collaboration of a number of agencies.
“Organizations should be able to negotiate amongst themselves to deliver better services at the conceptual level, then have as few technical impediments to the exchange of information as possible.”
Finding the Big Stick
From portals to procurement, citizen relationship management to centrally integrated service delivery, the goals and ambitions of e-government are clear. However, as is invariably the case, it is execution that poses the greatest challenge to ensuring that ITF does not end up as a technical footnote in the rubbish bin of Australia’s e-government history.
Australia is not the only country owning up to the challenge of delivering interoperability to fulfil the promise of e-government. Despite a concerted top-down push to promote intra-governmental interoperability within the UK, for example, that country’s e-Government Interoperability Framework (e-GIF) remains a work in progress. Aiming to promote uptake of the recently released version 5, e-GIF, managed by the UK Office of the e-Envoy, has been mandated for IT procurement and inter-agency joint projects.
However, e-GIF’s practical employment has faced internal challenges as individual agencies work to apply its technical prescriptions — which include XML and XSL (eXtensible Stylesheet Language) for data integration and presentation — to their ongoing business problems. To improve understanding and utilization of e-GIF’s guidance, the UK government established an information clearing house called UK GovTalk (www.govtalk.gov.uk) around which public sector agencies are uniting to develop best practice examples and function-specific implementations. GovTalk’s body of knowledge approach has most likely guided attempts to encapsulate governmental knowledge around Australia’s own ITF and its current ongoing review.
The UK may have mandated interoperability compliance, but government agencies in the US are working to promote interoperability by hitting departments where it hurts. Recognizing the unlikelihood that massive top-tier departments could be coaxed into following the country’s Federal Enterprise Architecture (FEA, at http://feapmo.gov/fea.asp), the Office of Management and Budget has begun to tie funding decisions to departments’ adoption of the guidelines. Proposals following FEA are given favourable treatment compared with those that do not, providing a financial incentive designed to get even the biggest departments into philosophical accord with the architects of FEA.
Whether either approach will work in Australia is a question of some importance. NOIE, the body that has directed the overall effort to develop and promote ITF, has long been criticized for lacking the kind of enforcement power and clout necessary to make either legislative mandate or financial incentives particularly effective. Furthermore, NOIE was thrown into disarray in April this year after the organization was terminated and its responsibilities split between a new Office for the Information Economy (OIE, a subunit of the Department of Communications, Information Technology and the Arts) and the new Australian Government Information Management Office (AGIMO).
OIE took over NOIE’s policy, research and programs work, while AGIMO’s responsibilities centre on IT and Internet projects applying across multiple agencies. Given that interoperability efforts such as ITF fall on both sides of the fence, the dissolution of NOIE could potentially complicate efforts to document and advocate ITF throughout the government community.
That does not bode well for those keen to push government departments into the new world of interoperability. “The challenge for government is to say where and how we can put together the different layers of government in a way that makes sense — and across these, what makes sense in terms of standard architectures,” says Gartner research director Steve Bittinger.
“If you’re in this more federated style of organization, interoperability frameworks are as much as you can do. If you want dramatic change, you have to look at who has the money, who has the power, and how you’re going to architect the collaboration that can occur through these frameworks.”
Learning Through Doing
In the short term, government CIOs will have to continue pursuing interoperability initiatives through sporadic and individualized projects that are regularly reviewed and assessed in terms of compliance with evolving standards like ITF. Given the myriad policies managing interoperability between departments, the most significant interoperability successes may well continue to be those negotiated on an individual basis between departments, as in the past; more open interoperability frameworks are still some way down the road.
For now, it may be more productive for individual departments to explore interoperability options that extend vertically, through all three layers of government, rather than necessarily focusing on horizontal interoperability between government departments. This may require bridges between federal interoperability guidelines like the ITF, and interoperability at other levels — for example, through the LGA-sponsored Local Government Interoperability Framework (LGIF) — but it can pay great dividends if appropriately managed.
One example of vertical interoperability paying dividends can be seen in the ISD (Integrated Service Delivery) Framework Project, a pilot testing the ability of Centrelink, the Western Australian government and local service delivery agencies to implement consistent multi-level policies managing information exchange.
Focused on Centrelink’s Customer Confirmation initiative, the ISD effort involved the establishment of an ISD “triangle” representing the conceptually cyclical structure of the interoperability process. This allowed local government service agencies to immediately confirm citizens’ eligibility for a variety of services based on information sourced directly from Centrelink.
With technical linkages in place, the triangular approach encourages sharing of knowledge, experience and policy formulation between involved parties. Ultimately, expanding such vertically applicable arrangements under the auspices of a revised ITF will provide guidance as to how other government departments can finally reap the benefits of a consistent, far-reaching interoperability platform.
Ultimately, the process of guiding interoperability will be an iterative one, requiring the work of everyone from DSTC researchers to CIOs, ministers and departmental managers across all levels of government.
“The best we can do is to try and give good ideas a run, to expose the economic benefits of these things, and raise the level of discussion to where there are compelling economic reasons for this new style of interaction,” says DSTC’s Wood. “When you sell it as a win-win that increases collaborative power, you get a sum that’s more than the parts.”
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.