A socio-technical framework for cyberinfrastructure design

As the concept of cyberinfrastructure (CI) matures people question whether it is something that can be designed or built. These doubts may stem from considering design as a time-limited process or the lack of a framework in which to view, integrate, and compare knowledge relevant to e-Science applications. This paper addresses these issues by identifying the components of a preliminary framework for CI design where design is viewed as an ongoing and long-term process. The framework is necessarily a socio-technical one because CIs interact with people, technologies, and other infrastructures. While there has been a commendable focus on science drivers and user needs in many CI efforts, more than this is needed if CIs are to adequately balance the needs of all users and other stakeholders. I argue that a framework for CI design should consider the following elements: maturity and complexity of the CI; users' capabilities, expectations, and perceived need for CI; research problems and their relation to computing; hardware, software, and associated technologies produced by others; the role and influence of all stakeholders; and mechanisms for coordination and feedback. This paper has two purposes: to introduce the framework and to communicate a new way to consider CI design.