Analysis: Collaborating With Documents, Managing With XML

To solve complex problems, sometimes it does take a rocket scientist. Sometimes, in fact, it takes a whole bunch of them.

That s what the National Academy of Sciences had in mind when it formed the Universities Space Research Association USRA- in 1968. Designed to foster cooperation between universities with graduate programs in space sciences or technology, USRA helps the National Aeronautics and Space Administration NASA- conduct technology and project research, and it now counts more than 95 colleges, universities and nonprofit research institutes in the U.S., Canada, Europe and Israel as members of a supporting research consortium. USRA is all about sharing information, helping thousands of engineers and scientists collaborate more effectively on research and development projects. But when it came to project management, USRA s technology had been anything but effective. Project managers would manually copy and paste structured financial data as well as content associated with project and task plans from Word documents, PowerPoint presentations and Excel spreadsheets into master documents and spreadsheets required for project management. An audit of one project that involved 70 separate project plans and more than 500 scheduling milestones revealed a 40-percent transcription error rate, underscoring the difficulty of manually tracking and synchronizing content versions and updates. USRA s goal was more effective reuse of source-document content in various master project management and performance management reports and registries. To ensure that projects are on track and on budget, these master documents and registries are compared against NASA s SAP financial management system and separate Oracle databases used to track technical projects. In 2004, USRA and NASA worked together to create a project management application that would synchronize data and content from templated Word, Excel and PowerPoint source files with the management reports and databases. The key was creating a system that could create XML-based, componentized renditions of documents together with searchable abstracts and context. The result was an extensible database overlay that runs on most industry-standard relational databases and that has since been commercialized in Xerox DocuShare CPX. In contrast to purely XML-based content management environments, the beauty of the CPX approach is that it manages XML renditions of content alongside original source documents, so collaborators can continue to use conventional documents and authoring tools. The key feature we wanted was simplicity, explains Dr. David Bell, director of the USRA s Research Institute for Advanced Computer Science RIACS-. We wanted to reduce the transcription errors, but to do so in a way that would allow users to use standard authoring tools they were familiar with. What s more, CPX is schemaless, so it doesn t require complex and time-consuming mapping of content. Instead, pluggable converters for popular file formats such as Word, Excel and PowerPoint create XML renditions of content along with access control rules, metadata and contextual analysis of the original documents. The database administrative support requirements for the system are very small, Bell says. Whatever XML files you drop in, it indexes them and they re queriable. The first two applications created were for financial tracking and performance measurement. Rather than copying and pasting, users now click on a button added to the document templates to either import content from or upload content to CPX. Serving as an XML-based hub that manages and stores these components, CPX tracks and synchronizes the most up-to-date information in these applications - including financial data, scheduling milestones and textual descriptions of facilities and risks - among source documents, the CPX database and core systems. With the first two applications still in beta-stage, the benefits of the project are still anecdotal, but the manual aggregation and update processes have been automated and the 40-percent transcription error rate has been eliminated. As we ve developed these integrated applications, there s a benefit in the transparency that results, Bell says. Information is coming in an unaltered way directly from the project managers and their documents. Project managers in turn get to see how the information from their projects will be presented in various management reports and other formats because the [management reports and registries] are available to them as well. And for their part, project managers appreciate no longer having to copy and paste content from one project artifact to another. Managers get a lot of requests to report information, and this has reduced their workload significantly, Bell says. USRA has not estimated an ROI for the project or the cost of developing the extensible database. As commercialized by Xerox, a 100-seat system of DocuShare CPX starts at $45,000, and it s being aimed at semistructured content management challenges such as contract management, legal discovery, research and incident reporting applications. Source: www.intelligententerprise.coma>