Gateways Unite! Findings from the Sustainability Blueprint Factory

XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE Gateways Unite! Findings from the Sustainability Blueprint Factory Nancy Maron BlueSky to BluePrint, LLC Yonkers, NY USA [email protected] Val McNutt BlueSky to BluePrint, LLC Claire Stirm San Diego Supercomputer Center San Diego, CA [email protected] Abstract— In 2023, the Center of Excellence to Extend Access, Expand the Community, and Exemplify Good Practices for CI through Science Gateways (SGX3) launched a series of Blueprint Factories, research studies seeking community input to address key challenges facing the scientific research community. The Sustainability Blueprint Factory was designed to investigate the challenges to the long-term viability and ongoing success of science gateways. The final report, published in June 2025, outlines five persistent challenges, shares examples of promising approaches, and indicates further exploration and action to create a robust, valuable, and sustainable online environment for scholarship. This conference paper offers a summary of the report’s findings. Keywords—sustainability, science gateway, business strategy. I. INTRODUCTION Over the past two decades the work of scholarship and education—research, writing, teaching, publishing—has slowly but inexorably continued to migrate online, morphing from paper to digital formats, from in-person to virtual workspaces. Few aspects of scholarship and teaching are now conducted without support in some digital form. Science gateways are one manifestation of this ground shift. These online interfaces give researchers, educators, and students easy access to shared resources that are otherwise inaccessible or unaffordable for a large segment of the scientific community. Among the resources gateways make widely available are high-performance computing capacity, data collections, scientific instruments and other analytical tools. As these science gateways came online, it became clear that fiscal and human resources needed to support these efforts would be significant. The rise of science gateways has mirrored similar efforts taking place in the humanities and social sciences, in library collections, and in the world of scholarly publishing: using technology to widely share the fruits of research and scholarship. While the specific content and tools may be quite different from the data sets and computational work of early science gateways, the underlying similarities were undeniable: large-scale web-based access for the community, often hosting large amounts of scholar-created content, aggregated online in order to more easily provide wide access to a global audience of scholars and students, an act of heroic sharing of work that in decades past would have been greatly limited in its scope by the constraints of local geographies, travel budgets, and physical or monetary barriers to access. As science gateway creators and their cousins in the humanities and social sciences soon found, building a gateway is costly and complex. The issue of “sustainability”—how to build science gateways and also digital humanities projects, library collections, and repositories that continue to be valuable and have reliable and recurring sources of support—has been a central challenge for each new community of practice as it has entered the digital realm. These innovative, community-facing digital initiatives share: • A focus on supporting knowledge creation and sharing • A mission to facilitate sharing of high-quality content, data, models, tools, and software with researchers and students (and sometimes the public) • Technical cyberinfrastructure, hosting hardware, storage, and tools that, once built, require ongoing support • Maintenance, and upgrades for the benefit of the current needs of the community • Challenges in finding new funding sources beyond the original start-up grants • An expectation of participation and interaction with a community Science gateways of all types looking to grow and thrive often struggle to: • Develop a reliable and continuous variety of revenue streams, whether due to a reluctance or inability to engage in revenue-generation, still sometimes seen as at odds with a mission-based project • Transition from experiment to reliable service and then navigate the change needed when a small group of academic researchers must become a customerfacing, customer-focused operation • Adjust to prioritizing the needs of the evolving community over one’s own preferences • Secure the expertise needed that may not be common to the core research team, such as project management, sales, marketing, and programming • Fully understand user/stakeholder needs and act on that knowledge to enhance the science gateway This report adopts a broad definition of “gateway” to include not just those research platforms that self-identify as “science gateways” but any large-scale, research-focused, digital initiative that has the intention of widely sharing cyberinfrastructure, tools, data, models, or content, in the service of facilitating knowledge creation and sharing. The reason for such a wide net is that this study is concerned with the challenges of supporting and sustaining these communityfocused digital initiatives. Whether STEM or humanitiesfocused, the challenges they face are remarkably similar. II. GOALS OF THIS PROJECT The Sustainability Blueprint Factory is a research project of the Center of Excellence to Extend Access, Expand the Community, and Exemplify Good Practices for CI through Science Gateways (SGX3), an NSF-funded, multi-institutional service organization. Since 2016, SGX3 and its precursor SGCI have worked with hundreds of gateways PIs and teams, observing first-hand the successes and challenges of creating and running a science gateway. The Sustainability Blueprint Factory is one of two Blueprint Factories currently underway. Its aim is to take stock of significant progress over the past 15 years, surface key challenges that still exist, and—perhaps most important— shine a light on promising approaches to addressing those persistent challenges. Today the notion of science gateway has expanded beyond its original remit to provide access to high performance computing resources, and this report embraces that broader view. This report includes examples from the worlds of STEM, humanities, and social sciences. Some of the innovative digital projects highlighted here self-identify as “science gateways” and others have leaders who might not recognize the term. We describe challenges facing large-scale, research-focused, digital initiatives that have the intention of widely sharing cyberinfrastructure, tools, data, models, learning objects, or images in the service of facilitating knowledge creation and sharing. If there is one key message underlying this exploration, it is that the challenges facing these seemingly disparate digital initiatives are so similar as to beg the question: Might there be bigger, bolder, more unified approaches to creating ways to support this still-new landscape of digital scholarship? The report includes five sections, each dedicated to a major challenge. Each section begins with a review of the challenge and progress made, offers examples of these approaches in action, and closes with a list of recommendations for future actions. III. METHODS The study is based on an extensive literature review of works relating to sustainability of digital initiatives. A first round of expert interviews and focus groups helped identify some key themes. These early interviews also helped surface a range of “approaches” to addressing these challenges. Subsequent interviews with people who were directly knowledgeable about these approaches helped to round out the picture. In total, the research team conducted over 35 interviews with gateway PIs, directors of units responsible for supporting gateways and/or software development for gateways, funders who support digital initiatives, and others deeply knowledgeable about the field. IV. KEY CHALLENGES AND PROMISING APPROACHES The Final Report identifies these major areas of persistent challenge, and highlights areas of progress. A. Too Many Gateways, trying to do it all Challenge: Many participants noted that academic independence and innovation, so highly valued in the academy, can act against building gateways that are easy to sustain. Approach: Borrow, don’t build. The rise of platform solutions has helped a great deal in certain spaces, from offering hosting for gateways (HUBZero, QUBES), compute capacity and tools (Galaxy, Open OnDemand), publishing platforms (Manifold, Fulcrum). Full hosting solutions can work but may be costly and some PIs prefer further customization. Approach: Convergence and “coopetition”. More than one funder reported actively seeking greater collaboration among grantees. At NIH, this took the form of a structured program called GREI for data repository leads, meant to encourage the leaders of these repositories to develop more interoperable systems to facilitate information sharing. B. Institutional Support Challenge: Institutions of higher learning, the home of many gateway PIs, have done a great deal over the last decades to support their innovators. One way they have done this is to create support centers. Digital Humanities, Digital Scholarship. More recently Open Source Program Offices are among the latest efforts to support faculty and students in this work. Approach: Campus Support Hubs. In the 2010’s many campuses created digital humanities centers, or digital scholarship centers, with the aim of providing training and support to those building digital work. More recently, we have seen the rise of the Open Source Project Office (OSPO), with the aim of providing expertise and training to those wishing to create gateways and other digital works using open source code. C. Value Proposition Challenge: At the heart of the success of any one gateway is the question: how does it provide value to its users? Campus administrators and funders we spoke with reported that PIs still struggle to pinpoint what makes their gateway truly distinctive and valuable to its target audience. Approach: Placing user needs first. One approach is to make sure the community of practice is deeply behind new gateways. Examples of gateways that have had long-term success by emerging from a community include the Archaeology Data Service (created in 1996) and The Protein Data Bank (1971). Approach: Training and Consulting. Several significant efforts have emerged to help PIs learn the skills needed to develop stronger value propositions for their gateway. The NSF-funded I-Corps program does this by emphasizing the importance of customer discovery work. The SGX3 Center of Excellence has also offered training to PIs in the form of consulting services and the Focus Week offering, a week-long intensive designed to coach PIs and their teams in executing and interpreting user-facing research. Approach: Standardizing impact metrics. It can be difficult to know if a gateway is succeeding in having its desired impact if the methods for measuring impact are still very much a work in progress. That said, pilots are underway to help develop and standardize impact measures. MakeDataCount.org seeks to measure how open datasets are accessed or utilized, while the CHAOSS project seeks to better understand open source community health on a global scale. D. Sustainable Funding Sources Challenge: Most PIs, when asked about greatest challenges, point to scarcity of funding. (We might argue that the better refined and executed the value proposition, the easier it is to locate and obtain funding…) That said, having funders seed new gateways, and then hope that they will become selfsustaining is wishful at best. Approach: Collaborative funding models. In the last decade, we have observed several innovative funding mechanisms being piloted, an effort to reframe the logic of how to fund critical infrastructure for the long term. These include Invest in Open Infrastructure (IOI) and the Canadian-based Sustainability Coalition for Open Science Services (SCOSS). Approach: Funder sustainability tracks. While the grant funding landscape today is uncertain at best, until very recently, federal funders with significant investments in science gateways were making available funding tracks specifically to support gateways in either winding down a project that was coming to a close; or funding to support creation or implementation of a sustainability strategy. Seeing “sustainability” as a set of activities worthy of investment, and requiring expertise. E. Leadership, Staffing, Governance Challenge: Given funding scarcity, it is not surprising that paid professional staff is often in short supply on gateways, even those in existence for years. It is still not uncommon to see a gateway staffed by an executive director assigned at less than .5 FTE; or a significant project with lots of workers, but only 2-3 paid staffers. Approach: Professionalization and the rise of the RSE. Organizations like the Software Sustainability Institute have helped to define and legitimize this new form of labor that is so central to creating and maintaining science gateways. Approach: Culture of support. Acknowledging that the world of developers can seem (or be) quite inaccessible in some ways - work styles and culture, in particular, Openscapes seeks to encourage more people to engage in building technology, by establishing a shared language, destigmatizing asking questions, and openly working across colleagues to establish shared norms for how information is shared. These processes have had measurable results, accelerating the ability of a team to finish projects. Approach: Rise of Consortia. In recent years, consortia have emerged as a way to support new gateway-like projects. With creative thinking from groups like the Stakeholder Alignment Collaborative, perhaps new models will emerge that bring together partners, using governance models that are better able to share the burden of support of science gateways. V. LOOKING AHEAD The Sustainability Blueprint Factory study has pinpointed areas where significant work has taken place over the last 15 years and indeed many signs of progress are visible. But sustainability challenges remain despite a range of different approaches taken by PIs, institutions, and funders to offer the support needed. Going forward, we see the need for: • Renewed focus on deeper collaborations and shared solutions, so that gateway PIs are not obliged to do all of the jobs needed to run a gateway. Re-use of existing solutions and integration of existing solutions could also help focus attention among a community of users. • Finding community where it already exists. Taking the lead from some great older gateways that emerged from within a community of practice, encourage funders and gateway leaders to demonstrate demand before making key and costly decisions about new gateways. • Who pays? Continue to ask what people or what organizations ought to be supporting these gateways? Is this part of the mission of the university as a center of learning? Is it the responsibility of a scholarly community to invest in its (open) platforms, data, and content? To what extent are certain gateways (or elements of them) significant enough to be considered public goods and worthy of ongoing public or community support? We hope the five major challenges discussed in this paper and the promising approaches to solving them can serve as the basis for work sessions with groups of stakeholders, to define actual blueprints for taking these ideas even farther or building them out at larger scale. ACKNOWLEDGMENT This paper is a summary of the Final Report of the Sustainability Blueprint Factory, a research study conducted on behalf of the SGX3 Center of Excellence to Extend Access, Expand the Community, and Exemplify Good Practices for CI through Science Gateways, funded by the National Science Foundation. REFERENCES [Right now the paper is not footnoted. I can certainly add back in references throughout the text if desired.] [1] [ to come]