Managing CI Centers: An Agenda for Organizational Scholarship and Cyberinfrastructure Innovation

Abstract

We present the following findings to a panel of CI center executives, senior organizational scholars, and CI policymakers in an NSF-sponsored “Managing CI Centers” workshop, and asked them to critically scrutinize and comment on these findings.

(1) CI Centers as Cyberinfrastructure “Stewards”: To date, the bulk of studies into organizing around cyberinfrastructure focus on project teams. Our research extends this work by situating these project teams within their organizational contexts, and focusing on an essential role played by CI centers, namely, that of cyberinfrastructure steward. This form of stewardship entails combining a service ethos with a responsibility for sustaining elements of a digital infrastructure over the long term. This is accomplished through the coordination of knowledge resources (e.g., people, hardware, software, documents) across projects. As stewards of cyberinfrastructure, CI centers are positioned centrally within contemporary scientific and technological networks. Accordingly, we also propose that CI centers, through their position at a cyberinfrastructure network’s “structural fold," effectively enable cyberinfrastructure innovation over time.

(2) CI Center Leaders as Entrepreneurs: Leadership in CI centers is usually associated with project management. Our research has found that, in practice, CI center leadership more accurately resembles entrepreneurship. CI center leaders come from a variety of backgrounds with diverse dispositions. We find that the attention of CI center leaders is a critical element of cyberinfrastructure evolution.

(3) Resource scarcity and Cyberinfrastructure Innovation: The single most important concern of CI center leaders involves access to the appropriate resources - both funding and human resources - to enable them to innovate. Although, in their view, more resources enable greater innovation - our findings indicate that this is not the whole picture. We find that the interplay of resource abundance and resource scarcity result in a sort of “generative tension” that fosters innovation.

(4) Scientific Software Ecosystems: Software is a critical element of cyberinfrastructure. Our research has examined the production of software in science and compared it with both commercial software production and production in the open-source software communities. Software development in CI centers, whether for internal or external use, largely takes place in relatively small, cohesive groups known as “software projects”. Increasingly, though, CI centers are looking to participate in “software communities” either by attracting outside contributors or, less frequently, by contributing to existing software communities. Yet our work also shows a third role, as yet relatively unexplored in science: the software ecosystem steward, working at the broader community level to establish incentives, collaboration infrastructure, and governance for coordinated development.

(5) Assessing CI Center Impact: The impact of CI centers is often invisible but it is broad and far-reaching - resulting in everything from technology innovation (like key elements of the Internet) to groundbreaking science in multiple disciplines. However, because of the invisibility of infrastructure, assessing their contributions is one of the critical challenges for CI centers and policymakers. While difficult, fair assessment is imperative when navigating various institutional arrangements, acquiring resources, and comparing potential cyberinfrastructure investments that will shape future scientific patterns of action.

Keywords: CI Centers, Cyberinfrastructure, Managing, Management, Organization Science, High-Performance Computing