Last month, I participated virtually in the Joint Roadmap for Open Science Tools (JROST) workshop as 86 individuals from 58 different organizations gathered in Berkeley on August 27-28 to explore the growing category of open source scholarly workflow tools, to compare notes, and to identify areas of cooperation and integration.The event program includes all the background information and presentations so I’ll try to highlight only a few key takeaways, noting that I found the entire program useful and thought-provoking.  JROST is an initiative worth watching because it seems to have the potential to meld together a coherent open ecosystem that could compete effectively with the commercial alternatives my colleague Roger Schonfeld has written about.

Greatly benefiting from the vision and diligence of Dan Whaley, founder and CEO of Hypothesis, the JROST initiative began earlier this year as a collaboration to form a common vision to support open science research workflows and to coordinate work across the community of open science projects spearheaded by non-profit organizations. The academic tool marketplace presents a range of applications designed  to support various stages of the research lifecycle, starting from the earliest steps of exploring an hypothesis to identifying collaborators to establishing preservation protocols and demonstrating impact. There is an increasing interest in the higher education community to understand the emergent marketplace and to identify tools that will offer solutions to challenges the academic community face. The initiative stemmed from a workshop associated with the FORCE2017 conference organized by Coko and Hypothesis and based on community interest in forming an alliance. The ultimate goal of the initiative is to consider the researcher experience in a more holistic way and pursue synergies to bring efficiencies and innovation into the scholarly workflow. To that end, developing user stories to understand how such tools are deployed by scientists is at the heart of the initiative.

The workshop began with an overview by Bianca Kramer and Jeroen Bosman of the open science tools ecosystem, which provided an excellent framework as the first day focused on looking at tools for discovery, analysis, writing, publication, outreach, and assessment. They presented hypothetical research workflows to compare a set of alternative products from for-profit and open science organizations. Based on their impressive work coined as 101 Innovations in Scholarly Communication, they have been systematically charting the scholarly landscape to explore the ways information is created, shared, and processed in academia. The second day’s opening presentation by Lettie Conrad was quite enlightening as it built on the framework described by Kramer and Bosman to visualize the current ecology of open science tools in order to illustrate areas where there is need for additional tool development. For example, although many open science journals platforms include content alerting or recommendations features, there are no tools across these platforms to keep researchers up to date on their topics of interest. She also pointed out that, even in areas where there is an abundance of tools, there might be shortcomings in usability, accessibility, or interoperability with other systems. She also pointed to some disciplinary discrepancies, underscoring the inevitable privileging of tool development in the biological and life sciences due to the higher interest in these fields by funding agencies.

There were a couple of dozen flash talks describing how specific initiatives fit into the open science ecosystem. The presentations demonstrated the burgeoning and versatile open source space with a range of tools that focus on various stages of the scholarly workflow and operate at different scales and organizational settings. One of the common threads throughout these talks was whether these tools have durable business models and can ensure service continuity.  Judiciously, the workshop included a significant emphasis on the sustainability challenges for nonprofit initiatives. The presentations and subsequent discussion on this topic were lively and conveyed the community’s recognition of the importance of considering long-term maintenance, development, and funding issues. The discussions also revealed a gap in open source software development in general; there is a need for best practices and guidelines for administration and fundraising. The announcement on September 7 that Authorea, which provides an online collaboration tool, was acquired by Atypon to join the Wiley family is a reminder of the fluid state of business models. The assertions that framed the session on funding the ecosystem summarize some of the key points:

  • The “ecosystem” is infrastructure.
  • Funding needs to be ongoing.
  • The end game is to reallocate funding now being spent on the closed ecosystem to the open ecosystem.
  • There is a need to create a community, therefore governance is as important as funding.
  • Funding with the largest possible players is best (governments versus individual institutions)

Prior to the workshop, the participants proposed a list of workflow topics and identified nine specific areas for deeper analysis through breakout sessions:

  • How can archives and repositories better integrate with each other and with the researcher workflow overall?
  • How can the community work together to improve publishing, submission, and editorial activities and integrate better with other open systems?
  • What are the researchers’ needs around of collecting, analyzing, computing, and storing research data, and how can we better meet them?
  • How can we open the workflow of peer review and at the same time look at ways of making the practice of peer review more transparent? How can annotation play a role?
  • How can we bring further efficiencies to the discovery, search, and access process to cope with information overload?
  • How can the tool creators for scholarly identity and identifiers work together more effectively?
  • How can citation management tools be integrated most effectively into researcher workflow and with the other tools needed in the writing and publication process?
  • How can we help researchers work together more effectively, as research is an collaborative undertaking?
  • How can various data tools integrate smoothly?

As I reflect on the workshop, what impressed me the most was the approach of software developers to this space as a public good and their genuine interest in sharing common practices and funding strategies.  As I attended virtually, I sensed the energy and collaborative spirit even through my computer monitor. The grand challenge ahead is balancing the community’s creativity in designing versatile user-centered open source platforms with a collective effort to develop models to ensure ongoing financial investment.  The open ecosystem needs to be prepared to compete with the commercial and closed systems. This requires that open source initiatives follow the proven best practices in marketing and customer relations, as the researchers are likely to select tools based on an application’s functionality and practicality, rather than their interest in supporting open science. Also essential is making an effort to take a balanced approach, so that we don’t experience tool gaps for different disciplines, especially those in the social sciences and humanities, due to the nature and availability of start-up funds.