Quite recently, the National Science Foundation rewarded Daniel Atkins for his unique contributions to high-performance computing and distributed knowledge architecture with a division of his own, the formidably named Office of Cyberinfrastructure. In his talk, Atkins outlines his ambitious goals for this new post.

In a detailed vignette, Atkins describes a research enterprise that set him on his current course of advancing “learning and discovery without geographic, temporal, cultural or institutional barriers.” During the 1980s, Atkins participated in a multidisciplinary, upper atmosphere study that enabled scientists to use telescopes and other instruments housed in a remote Greenland site from their home institutions. In pre-World Wide Web days, Atkins helped design a computer network that made it possible for researchers to share and display data, and gain real-time access to instruments and to “the entire spectrum of resources that a community of practice needs to do its work.” Over years, this “collaboratory” led to “dynamic workrooms” across many time zones, inspiring scientists who otherwise competed against each other and who hogged precious instruments to collaborate and pool resources. In addition, this science portal became accessible to many others, including college-age and younger students.

This was a defining experience for Atkins, who grasped that the collaboratory enabled a new kind of research, and with “appropriate scaffolding,” supported education and outreach efforts as well. “Multiple outcomes are possible if we do it right; we need to create an environment that will create more synergy between research, educational, and societal engagement,” says Atkins.

This is precisely what he’s after at NSF. Acknowledging that laying the foundations for a widespread, interconnected system of research communities will be expensive, Atkins notes that cyberinfrastructure (CI) is already emerging worldwide, and that at NSF, it will be “both the object and the means for R&D.” NSF will push forward on several fronts simultaneously: high performance computing, data analysis, virtual organizations, and learning and workforce development, hoping to apply CI “in transformative ways.” This is a daunting and humbling task, especially given the resources, says Atkins, and will depend on cooperation among many groups from the get go: “Anything that smacks of revolution has got to be a multi-stakeholder enterprise,” he says. But Atkins believes the “inherent complexity and multi-scale nature of frontier science challenges” demand a multidisciplinary, multi-institutional and frequently international approach that will depend on CI. He also believes part of the solution to our nation’s crisis in science and engineering education lies in encouraging public interest in new and inviting online science communities.