The national cyberinfrastructure (CI) ecosystem is essential to computational- and data-intensive research across all science and engineering (S&E) domains. The CI ecosystem is highly dynamic, driven by rapid advances in a wide range of technologies, increasing volumes of highly heterogeneous data, and escalating demands for CI resources and services by the research community. Innovations in CI are consequently key catalysts for new modes of discovery and play a critical role in ensuring U.S. leadership in science, engineering, economic competitiveness, and national security, consistent with NSF's mission. It is thus imperative that CI innovations become available, in a coordinated and sustainable manner, as part of the NSF-funded advanced CI ecosystem.
The Advanced Cyberinfrastructure Coordination Ecosystem of Support and Services (ACCESS) program provides an array of national-scale CI services to the S&E research community, including integrated coordination of users requests for computational and data resources; integration of resource providers systems; deployment of technical support; monitoring of system usage; user training; and communication and outreach to the CI and research communities. The ACCESS program is envisioned to include a process to enable breakthrough CI innovations of recognized value to researchers to be translated into production-quality sustained services that are deployed and made available to the user community from applicable NSF-funded resource providers. Such a process would also include methods to prioritize which innovations to translate to production services and to identify when these services should be taken out of service or replaced.
This CITAP solicitation is focused on the translation of innovative research CI software such as system software, libraries, application codes, and software enabling data services. NSF seeks proposals that aim to design, test, and subsequently operate a pathway service within the ACCESS program that manages and accelerates the translation of promising research CI software to production-quality services across the NSF advanced CI ecosystem in support of the NSF S&E research community. CITAP proposals are expected to create a new workflow process (represented notionally in Figure 1 below) within the ACCESS program that: (1) identifies novel CI software from diverse sources in a strongly community-informed way; (2) establishes an open and merit-based process for selecting and prioritizing/sequencing which of the identified innovations are of highest and most immediate value to users of the advanced CI ecosystem and can be feasibly translated to production level and made available for use by researchers using ACCESS resources; and (3) establishes an operational process that translates innovations into production services, including creation of partnerships where necessary to address each of the technical challenges and intellectual property (IP) considerations faced when integrating novel CI software within the advanced CI ecosystem.
Examples of translation challenges include (but are not limited to): awareness by potential users who could benefit from the technology; systems engineering and software deployment; an initial launch period of collaborative maintenance; anticipation of user support that will be necessary; engagement of people or possibly organizations needed to sustain and upgrade the software/service in response to changes in technology and/or feedback from users and communities served; and determination of when the relative use or value of the software has decreased to an extent that decommissioning is warranted. The pathway must also be designed to evolve as the types of systems and services that are part of the NSF-funded production advanced CI ecosystem evolve and expand.
Figure 1. Stepwise translation of innovative software into production services in NSFs advanced CI ecosystem, followed by optional collaborations for upgrades or spinoff enterprises.
CITAP proposals should assume the following three phases of development activities during the award period for the CITAP project: process design; pilot development and testing using CI translation use cases within the ACCESS program; and initial operations at scale. Advancement by the project to each subsequent phase will be based on successful associated milestone reviews that will assess the satisfaction of performance criteria defined for each phase that reflect the interests of key constituents, including S&E users, system administrators, and CI developers. During the initial operations phase, product launches will include a period of collaborative maintenance by the CITAP awardee and developers. Longer-term maintenance should also be planned to sustain and upgrade the technology.
Given the variety of sources of CI research and potential development partners, CITAP proposals are expected to promote the sustainability of the production software via collaboration with organizations or relevant NSF projects. Office of Advanced Cyberinfrastructure (OAC) programs highlighted below are among those that foster development of CI technologies that are candidates for accelerated translation to production. Programs in OAC and in NSFs new Directorate for Technology, Innovations and Partnerships (TIP) support communities that may be suited to sustain or upgrade the production CI technologies. CITAP proposals are welcome to also include or propose to recruit collaborative involvement of constituents beyond NSF programs, such as academic institutions, non-government organizations, industry, and other government agencies.
Thomas GulbransenProgram Director, CISE/OAC