The Future of E-infrastructures

78 Computing in Science & Engineering 1521-9615/15/$31.00 © 2015 IEEE Copublished by the IEEE CS and the AIP May/June 2015 T he search for the Higgs boson, the huge advances in genetics, the improvements in weather forecasting, the growing understanding of climate change, and many other recent and ongoing high-profile scientific research endeavors depend on collaboration among scientists. These collaborations in turn depend on hardware and software infrastructure, referred to as e-infrastructure in the UK and as cyberinfrastructure in the US (we use the terms interchangeably in this article). Large amounts of public money have been invested in e-infrastructure—in the UK alone, the e-Science program funded projects worth more than £213 million between 2001 and 2006 (www.epsrc.ac.uk/newsevents/pubs/rcukreview-of-e-science-2009-building-a-uk-foundation-for-thetransformative-enhancement-of-research-and-innovation/). After the program’s completion, the UK Engineering and Physical Sciences Research Council (EPSRC) consulted with the relevant community to reflect on how e-infrastructure might best be developed and deployed in the future. These deliberations resulted in three documents published in 2012: a strategy document and an action plan (both available from www.epsrc.ac.uk/research/ourportfolio/themes/ researchinfrastructure/subthemes/einfrastructure/software/), and a report on research software engineers (www.software. ac.uk/resources/get-speed). These documents were synthesized and expanded into an e-infrastructure roadmap (www.epsrc. ac.uk/research/ourportfolio/themes/researchinfrastructure/ subthemes/einfrastructure/strategy/roadmap/). This roadmap is intended to be used by both scientists and peer reviewers to check that projects being proposed for funding are consistent with the national strategy. In the US, the National Science Foundation (NSF) published similar documents in 2012, setting up a framework for developing and deploying cyberinfrastructure (www.nsf.gov/cise/aci/cif21/CIF21Vision 2012current.pdf ), as well as a vision and strategy for software for science, engineering, and education (www.nsf.gov/ publications/pub_summ.jsp?ods_key=nsf12113). Recent research has revealed that an e-collaboration’s success doesn’t depend solely on its hardware and software underpinning. Rather, the people involved and the cultural, social, and organizational contexts in which they work—the human infrastructure—play a significant role.1,2 Here, we focus on the human infrastructure involved in developing the underpinning software rather than on the scientific collaborations per se. We analyzed the funders’ documents from the perspective of this focus and in the light of recent empirical studies on e-infrastructure development. Throughout this article, we include illustrative examples from a four-year-long field study (2006 to 2010) of the development of a relatively small e-infrastructure software package, a laboratory information management system for the use of protein scientists that we call FLIMS (field study laboratory information management system).