The SAM development project will be managed as a single distributed team working to develop a single coordinated set of components. A variety of mechanisms will be used to coordinate with other projects to ensure maximal benefits to the SciDAC and general DOE communities. We will present the proposed scope and schedule of the project at the initial National Collaboratories meeting in September and take advantage of other such outreach opportunities during the initial months of the project. These meetings will serve as a vehicle for soliciting interest in SAM and in obtaining feedback for detailing and prioritizing developments. Preliminary design documents for services will be developed and distributed to collaborators for review. A SAM website, loosely modeled on those developed for the DOE2000 notebook projects, will be developed to provide access to software, documentation, and help resources. We will investigate possibilities for appointing specific project liaisons and creating or joining a cross-project group focused on the role of SAM and other middleware technologies in generalized collaborative problem solving environment / portal architecture.
The project is planned as a 5-year effort with base capabilities becoming available for beta use within 12-18 months of the project start. User feedback will be used to refine these components concurrent with the development of advanced capabilities. As the project progresses, the SAM team will work with current and prospective users to identify mechanisms for long-term support of SAM capabilities. In the final phase of the project, we will transfer the support and evolution of SAM to other sources, while we focus on identifying next-generation annotation challenges for DOE and investigating prototype solutions.
The SAM team will work closely with other SciDAC developers to move toward standard representations for software-generated metadata such as data pedigrees (e.g., experiment parameters, system description, input files, software/algorithm version used), summary information (e.g., low-resolution subsets, identified features), and relationships to other data (e.g. part of a project or parameter study). The SAM team will work with SciDAC end users to define and develop a set of basic annotation types and semantic relationships necessary to represent project plans, hypotheses and conclusions, ideas for follow-on experiments, meeting notes, etc. At both levels, the SAM architecture will be developed so these standards can evolve and be customized as needed by specific groups.
The SAM project team intends to leverage emerging resources for the formative and summative evaluation of collaboratory interactions being developed within the NSF sponsored "Science of Collaboratories" project at the University of Michigan. This project aims to capture the lessons learned in DOE, NSF, and NIH collaboratory efforts and to develop guidance for planning, building, measuring, and maintaining scientific collaboratories. Materials from this project regarding success metrics, software instrumentation, and evaluation procedures will be leveraged in our effort and opportunities for closer collaboration will be explored.
Software development will be managed through a single version control and bug tracking system available to developers at both sites. We will investigate the possibility of making the project source code directly available to coordinating projects through, for example, licensing the software as open-source and using third party Web/CVS hosting. Managing to the shared vision presented above will be important for the success of the project. Weekly teleconferences and ad hoc data conferencing will support the team in developing a detailed system design, discussing development issues, and integrating the testing of SAM interfaces and components. Similarly, close interactions with application and Collaboratory pilot projects regarding interface design, development schedules, and deployment tasks will be handled through a combination of email lists, Internet conferencing, and traditional meetings. Project decisions will be made by a consensus of the project PIs, guided by advice from the SciDAC community.
This project team will perform system software research, development, and incremental deployment of SAM broken into three phases:
Phase 1: (~15 months) will include architecture definition, requirements gathering from SciDAC and related projects, technology selection, and basic operational organization. We will begin to build basic SAM components and services according to the requirements and anticipate release of beta capabilities by the end of this phase. Key foci will be development of basic metadata services and a transitional DAV-based notebook capability, reusing DOE2000 technologies that can be delivered to collaboratory pilots.
July 2001
September 2001 (3 months)
December 2001 (6 months)
March 2002 (9 months)
July 2002 (1 year)
September 2002 (1 year, 3 months)
Phase 2: (~18 months) will focus on incremental improvement and expansion of SAM capabilities. As we deploy versions of SAM components we will begin to institute monitoring and feedback activities to understand how the system is being used and will use this formation to guide further developments. During this phase we expect to maintain the basic system and add to its capabilities. Key foci will be the implementation of semantic and notebook services.
December 2002 (1 year, 6 months)
July 2003 (2 years)
December 2003 (2 years, 6 months)
Phase 3: (~18 – 24 months) will begin the transition of SAM maintenance to other sources allowing for advanced research into emerging issues related to SAM’s goals of improving the utility and completeness of scientific records. Key foci will be the expanding the functionality of integration components and the SAM notebook interface. As users come to rely on SAM, we will seek to assure long-term support by fully transitioning SAM evolution to the open source community, and/or investigate other methods of support. As this occurs, the project will focus on identifying next-generation annotation challenges for DOE and investigating prototype solutions.
July 2004 (3 years)
July 2005 (4 years)
July 2006 (5 years)