NSF Award
1322305
The motivating vision behind "computational and data-enabled science and engineering" (CDS&E) is that the development, implementation, adoption, use and diffusion of advanced cyberinfrastructure (CI) will revolutionize research and education in the U.S., enhancing the nation's economic competitiveness and maintaining the nation's leadership in science and engineering discovery. However, the implementation and diffusion of CI-enabled science -- and especially the adoption of advanced CDS&E tools -- is not well understood. These next-generation technologies are dynamic to an unprecedented degree, in that they are user-driven, custom-made, produced on demand, and "permanently beta." Most importantly, they are being put to use while simultaneously being developed; and all this is occurring within CI-enabled virtual organizations (VOs), which constantly evolve along with the technologies. <br/><br/>This project examines the rise and fall of new computational tools within CI-enabled, virtually-organized science and engineering research. The project employs a grounded-theory approach, to analyze interview data about the emergence of advanced CDS&E tools, and to construct a theoretical framework for understanding dynamic innovation development, use, adoption, implementation, and diffusion within CDS&E VOs. The project's empirical focus will be the NSF-funded Extreme Science and Engineering Discovery Environment (XSEDE; https://www.xsede.org/), a nationwide network of distributed high-performance computing resources. Snowball recruitment will be used to identify domain scientists, computational technologists, and supercomputer center administrators (across e-science projects, institutions, and disciplines) who have experience with adopting, developing, and using computational tools within the XSEDE ecosystem. Interviews will seek to illuminate best practices and common pitfalls in the emergence of advanced computational tools and in the deployment of such tools by practicing CDS&E researchers.<br/><br/>Findings from this work will help NSF and other federal agencies to optimize existing and future CI investments. The research may also have implications for commercial and open-source technologies that share some of the dynamic, distributed feedback and network properties of advanced scientific CI -- promising insights that could enhance the development and diffusion of such technologies well beyond the academic science community.
