Our journey began with our Department of Defense research projects when we saw an opportunity to solve our data management challenges with XML (http://xml.sys-con.com/read/40411.htm). The journey continued with the evolution of that work as it applied to exposing legacy data sources as XML through data-oriented Web services (http://xml.sys-con.com/read/45527.htm). We continue to build upon this foundation by broadening our service-oriented architecture (SOA) with these XML-enabled yet disjoint systems as we look to Business Process Execution Language (BPEL) to orchestrate their complex interactions.

My primary focus over the last year has shifted from enabling legacy data sources and applications as services to orchestrating the business processes that encapsulate those services. The motivation behind this work is to bring automation and repeatability to decision-support environments, thereby allowing the decision makers to focus on the big picture. People naturally associate BPEL with business or corporate environments where time is money; however, there is a poignant case to be made for leveraging this technology within government agencies where return on investment is measured in human life rather than in dollars. I'll present examples that demonstrate how applying XML and Web service technologies coupled with BPEL enhances the ability of decision makers to quickly assess and respond to situations with time-critical requirements. Also included is an example of US Naval support in humanitarian aid and disaster relief operations such as the undertakings following the tsunami in Southeast Asia and the hurricanes in the Gulf Coast.

Brief Review of Previous Work
When we first began developing with XML we initially set out to achieve the separation of content from presentation. We published our data as XML using our XML Data Services engine and used XSL transforms to generate our portal-enabled user interfaces. It soon became apparent that the real value in what we were able to do with our data sources would be realized when the raw information rather than the visualization could be shared. Out of Web services emerged defining standards that dramatically increase system interoperability and data sharing. The Web services effort to achieve interoperability using standards and a loosely coupled integration model allows flexible composition of systems in a variety of domains.

Business Process Execution Language
Systems integration is composed of much more than data sharing or the ability to carry out simple exchanges between services. Web services conform to a stateless model of synchronous exchanges or uncorrelated asynchronous conversations that can be invoked using an interface described in a standard format, namely Web Services Description Language (WSDL). BPEL receives the full benefit of past research and development through broad vendor participation, but also builds on that foundation by embracing emerging standards and has become the industry standard for Web service orchestration. SOAs become viable cross-domain integration platforms, given the ability to model complex interactions as well as a business protocol that supports synchronous and asynchronous, long-running, stateful, machine-machine processes. Figure 1 depicts the composition of discrete services from diverse systems participating in a complex business process.

In addition to its obvious technical merits, BPEL provides additional value by bringing the organization's business processes and procedures to the forefront during development. BPEL delivers the visibility into an organization's methods of operation and the business rules they practice, formally describing and documenting their business processes. This formality and transparency gives rise to process engineering, whereby an organization is able to focus on and validate business practices. As with the services BPEL orchestrates, the instantiations of these processes are discrete, reusable components capable of being composed into larger business processes or executed outside of the context of a larger application.

BPEL Supports Time-Critical Decision Making
Industry and government organizations alike are faced daily with time-critical situational assessments that have far-reaching implications. The Department of Defense is continuously engaged in operational exercises and scenario development. The vast experience and knowledge gained through these activities transition to operational procedures, commonly referred to as preplanned responses. Likewise, corporate scandals led to increasing regulatory compliance outlining standard operating and accounting practices. These well-established procedures and precisely defined responsibilities lend themselves to process automation.

First we'll look at a simplified industry example of business process automation to introduce the value of BPEL in time-critical decision making. The example I'll use is that of an Internet loan brokerage firm. This firm's job is to match borrowers with financial institutions willing to lend money to that individual for a prescribed amount and purpose. The broker's motivation for employing BPEL is to downsize the required workforce through automation while simultaneously increasing the number of borrowers they pair with lenders. By succeeding in doing so, they will lower the cost of doing business and increase income by generating more fee-based revenue. As shown in Figure 2, the broker would author a BPEL composition with partner links for the borrower, x number of credit reporting bureaus, and y number of lending institutions.

The process would be initiated by a borrower providing required information by completing an electronic loan application. By using that information the BPEL is able to asynchronously request the applicant's credit score from all credit bureaus in parallel. The credit score responses coupled with the applicant's information is then able to be sent in parallel to all of, or based on the applicant's input to only certain, lending institutions via their partner links. Each lender will have its own proprietary business rules to determine whether or not to offer a loan and what terms that loan will carry. The broker then asynchronously receives the responses from the lenders, presenting any offers to the borrower. Even in this simple flow, we are able to see multiple examples of time-critical assessments. The broker's efficiency is greatly increased by enabling the information collected from the borrower to be communicated in a machine-machine fashion rather than requiring a person to read, interpret, and route the input to a loan officer within the lending institution. That machine-machine communication is also done in parallel to the lending institutions, thus greatly increasing the number of opportunities to match the borrower with a lender in the same amount of time. The parallel flow additionally prevents deadlock of the process by removing the dependency of a response by one lender before submitting the request to other lenders. The lenders themselves reap the benefit of participating in the broker's flow by being exposed to more applicants. Moreover, if the partner link that the lender exposed to the broker were itself a BPEL, the lender would be able to increase the number of applications it could respond to through automation. Finally, building on loosely coupled services allows partners to change their proprietary back-end processing without affecting the business process. If a lender were to institute new business rules that qualify more borrowers, that module could be deployed behind the scenes without exposing the details to the broker's flow. This flexibility and the extensibility to easily add partner links to additional financial institutions positions the broker for future success.

One of the lesser-known missions of the US military is global humanitarian aid and disaster relief. The example below will focus on US Naval components where missions typically entail transporting and delivering tons of supplies, including food, water, and medicine, as well as hosting nongovernmental organizations such as the Red Cross and by serving as floating hospitals for some of the world's best physicians.

Figure 3 depicts the high-level flow between the following seven partner links: a Command Support System that manages facilities and supply status, a Personnel Management System that tracks among other things job descriptions and deployment locations, a Joint Medical System that maintains personnel medical records, a Weather System that provides current and forecast meteorological and oceanographic conditions, an Integrated Materials System that maintains locations and types of industrial materials, a Hazardous Plume Analysis System that supplies the capability to simulate effects of hazardous material releases, and finally, a Geospatial System partner link for map-based visualization and collaboration.